ossim-developer — Software development list for OSSIM

Re: [OSSIM] Orthorectification results: Ossim vs. Summit

[Guillaume D. <dro...@gm...>]

Forgot to ask:

Is there a document that defines what all the variables in Ossim geometry
(.geom) files mean (e.g. adjustment_0.*, distortion.*)?

Cheers.

2018-05-15 17:46 GMT-04:00 Guillaume Drolet <dro...@gm...>:

> Oscar,
>
> The 1 arcsec DEM has been smoothed using a gaussian filter, hence its
> smooth look compared with the 100m DEM. The elevation profiles are for a
> diagonal that goes from the lower left corner of the image to its upper
> right corner.
>
> Re Summit, as far as I know, Summit is never involved in the
> aerotriangulation process. Aerotriangulation is performed as one of the
> first steps prior to calculating the affine transformation matrices in DVP
> (i.e. the variables in the PAR file).
>
> Also, the TIN used in Summit to orthorectify the image (the one thats
> "almost" perfectly lined up with the road network and the Google Earth
> image) has much less points than the TIN referenced in the DVP PAR file
> (i.e. the .tgl file).
>
> At this point, I suspect that either 1) Summit uses more information from
> the DVP PAR file compared with Ossim, e.g. affine transformation matrices,
> or 2) Ossim derives transformation matrix coefficients using the Applanix
> sensor model, which may not be fully compatible with the UltraCamD camera.
>
> In any case, I will do as you suggested and try to gather more details
> about the complete workchain from image acquisition to creation of the DVP
> PAR file. I may post more information in this thread if need be.
>
> Thanks a lot for your help,
>
> Guillaume
>
> 2018-05-14 16:03 GMT-04:00 Oscar Kramer <osc...@ya...>:
>
>> Guillaume,
>>
>> RE the DEMS: weird that the 1 arcsec (~30m) is smoother than the 100m
>> DEM. I don't know where in that plot your image falls, but there are areas
>> of significant bias that could explain the scale difference observed.
>>
>> RE the Summit code: I have a feeling that your testing is incestuous.
>> Perhaps you are using data that was triangulated with Summit using the
>> reference image as ground truth, and then using the same model to project
>> to the same ground truth. There is no way to explain the perfect alignment
>> to the reference image otherwise. It would be a freak of statistical nature
>> to have it line up out of the box like that. Check the chain of custody of
>> that data set in your organization to see exactly what processing was done
>> to it.
>>
>> Sorry I can't be of more help. I don't have the time to dive into a
>> difficult problem like this.
>>
>> Good luck,
>> Oscar
>>
>> On Monday, May 14, 2018, 1:33:19 PM EDT, Guillaume Drolet <
>> dro...@gm...> wrote:
>>
>>
>> Salut Oscar, thanks for your help!
>>
>> I added more info in your email below. I am not a photogrammetry expert
>> so I surely do not understand all the details behind the orthorectification
>> process.
>>
>> G
>>
>> 2018-05-14 10:19 GMT-04:00 Oscar Kramer <osc...@ya...>:
>>
>> Salut Guillaume,
>>
>> Can you give us a little bit of background to help you solve the problem?
>> I am not familiar with DVP, PAR nor Summit.
>>
>>
>> DVP is a photogrammetry software used, among other things, to build
>> stereoscopic models for pairs of aerial photographs. I don't know if it
>> also performs aero-triangulation. It is now discontinued and was replaced,
>> in my organization, with Summit (http://www.datem.com/summit-evolution/).
>>
>> PAR files are model files created by DVP and contain interior and
>> exterior orientation parameters. Summit can orthorectify an aerial
>> photograph using only the raw image, the corresponding DVP model file
>> (.par), and a triangular interpolated network (TIN) file. I don't know if
>> it can also use a DEM since I don't work with Summit.
>>
>> In case you didn't have a look at the PAR file I put in Dropbox, here's
>> the content of it:
>>
>> .GLOBAL
>> $DATE 03-03-2008
>> $TIME 15:56:18
>> $DESCR OPK -> DVP ...
>> $TYPE OPK
>> $FSCALE00 2.50000000000000e+004
>> $FILE00 M:\Dvp\Images\JD07137\opt\rgb\JD07137_12007.tif
>> $VERSION 5.5
>> .DESCR
>> .INTERIOR
>> $FCAM00 C:\Projet\G1918\UltraCamD.cam
>> $PARAFFINE00 9.00000000000000e-003 1.55854062294791e-018
>> -5.15700000000000e+001 8.13151629364128e-020 -9.00000000000000e-003
>> 3.34800000000000e+001
>> $PARINVAFF00 1.11111111111111e+002 1.92412422586162e-014
>> 5.73000000000000e+003 -1.01965198758660e-014 -1.11111111111111e+002
>> 3.72000000000000e+003
>> $PARM00 1 -5.15700000000000e+001 3.34800000000000e+001
>> 0.00000000000000e+000 0.00000000000000e+000 1.42108547152020e-014
>> -7.10542735760100e-015 1.42108547152020e-014 -1.42108547152020e-014
>> $PARM00 1 -5.15700000000000e+001 3.34800000000000e+001
>> 0.00000000000000e+000 0.00000000000000e+000 1.42108547152020e-014
>> -7.10542735760100e-015 1.42108547152020e-014 -1.42108547152020e-014
>> $PARM00 2 5.19300000000000e+001 3.34800000000000e+001
>> 1.15000000000000e+004 0.00000000000000e+000 7.10542735760100e-015
>> 7.10542735760100e-015 0.00000000000000e+000 -1.42108547152020e-014
>> $PARM00 3 5.19300000000000e+001 -3.40200000000000e+001
>> 1.15000000000000e+004 7.50000000000000e+003 -7.10542735760100e-015
>> 1.42108547152020e-014 0.00000000000000e+000 0.00000000000000e+000
>> $PARM00 4 -5.15700000000000e+001 -3.40200000000000e+001
>> 0.00000000000000e+000 7.50000000000000e+003 -7.10542735760100e-015
>> 0.00000000000000e+000 1.42108547152020e-014 0.00000000000000e+000
>> $DATE 03-03-2008
>> $TIME 15:56:18
>> $FOC00 1.01400000000000e+002
>> .RELATIVE
>> .ABSOLUE
>> .RELEVEMENT
>> .EXTERNE
>> .CORRECTION
>> .CALIBRAGE
>> .TIN
>> $TGLNAME M:\Dvp\Mne\JD07137\TGL\JD07137_12007.tgl
>> .INDEX
>> .LISTMODELES
>> .CORRECTION
>> .CAMCOR00
>> $NAME UltraCamD
>> $DATE
>> $TYPE 1
>> $FOCAL 1.01400000000000e+002
>> $PPA 0.00000000000000e+000 0.00000000000000e+000
>> $PPS 0.00000000000000e+000 0.00000000000000e+000
>> $XP_OFFSET 0.00000000000000e+000
>> $YP_OFFSET 0.00000000000000e+000
>> $PIXELSIZE 9.00000000000000e+000
>> $DISTORTION_TYPE 0
>> .EXTERIOR
>> $DATE 03-03-2008
>> $TIME 15:56:18
>> $XYZ00 2.62237954000000e+005 5.40876598900000e+006 2.51442300000000e+003
>> $OPK00 -2.24520000000000e-002 -2.24683000000000e+000 8.90675920000000e+001
>> .END
>>
>> I obtained a document explaining part of what this file contains. Here
>> are parameter-specific informations:
>>
>> $PARAFFINE00: contains the affine transformation parameters (a to f) to
>> go from image coordinates to photo coordinates. Transformation equations
>> are as follows:
>> X_photo = a * X_image + b * Y_image + c
>> Y_photo = d * X_image + e * Y_image + f
>>
>> $PARINVAFF0: : contains the inverse affine transformation parameters (A
>> to F) to go from photo coordinates to image coordinates. Transformation
>> equations are as:
>> X_image = A * X_photo + B * Y_photo + C
>> Y_image = D * X_photo + E * Y_photo + F
>>
>> In DVP, origin of image coordinates system is upper left corner.
>>
>> $XYZ00, $OPK00: exterior orientation parameters. The order of rotations
>> in DVP is Omege, Phi and Kappa.
>>
>> $TGLNAME: this is a DVP generated TIN file. It's a binary file I can't
>> read with my tools. However, I have a shapefile version of it (elevation
>> points).
>>
>> It looks like DVP also used information from the file UltraCamD.cam
>> (variable $FCAM00 above in the file), maybe to derive the coefficient of
>> the affine transformation  ($PARAFFINE00, $PARINVAFF0,  $PARM00 in the file
>> above)?
>>
>> I managed to get the UltraCamD.cam file used in DVP. Here's its content
>> if that can help:
>>
>> SYSTEMAP CAMC120
>>
>>  101.400
>>      -51.750 1       33.750 1
>>  -999999.999 0  -999999.999 0
>>       51.750 1       33.750 1
>>  -999999.999 0  -999999.999 0
>>  -999999.999 0  -999999.999 0
>>  -999999.999 0  -999999.999 0
>>      -51.750 1      -33.750 1
>>  -999999.999 0  -999999.999 0
>>       51.750 1      -33.750 1
>>   1   1
>>    -0.180     0.270
>>   4 0.0000000e+000 0.0000000e+000 0.0000000e+000 0.0000000e+000
>>
>>
>> I also have a calibration report for the camera but it's a pdf and
>> doesn't contain tabular data for distortion values, only plots. Here it is:
>> https://www.dropbox.com/s/hm7sr2tpy6c6f2g/Calib-Report_0034_
>> V70_short.pdf?dl=0
>>
>>
>>
>>
>> Also, what is the ground truth background you are comparing to on the
>> mosaics you provided?
>>
>>
>> It's a Google satellite map (OpenLayers QGIS plugin) as well as the road
>> network (white lines, 1:20K) produced by my organization (Ministry of
>> forests) and used for forest planning, etc.
>>
>>
>>
>> There are many things that can be an issue. You pointed out rotation
>> matrices. Perhaps the Applanix model if performing the rotations in a
>> different order (though I would expect a much bigger discrepancy in that
>> case). Also perhaps Summit can read DVP distortion parameters.
>>
>>
>>
>> Where would Summit read distortion values from? Could they be the
>> $PARM00  variables in the PAR file above? As I mentioned, Summit uses only
>> the information contained in the PAR file and a TIN to create the
>> orthoimage.
>>
>>
>> RE the 100m v 30 arcsec DEM results. I don't know what to make of that.
>> Indeed, the 100 m DEM seems to have a big scale error as if there was a
>> significant elevation bias. Did you compare the DEMs to see if there is in
>> fact a bias between the two?
>>
>>
>> I created a profile plot that compares the two DEMs (red: 1arc, blue:
>> 100m):
>>
>> https://www.dropbox.com/s/8widxe4lyx657wi/DEM_1arc_vs_100m.png?dl=0
>>
>> Note how the 1arc DEM has been smoothed. Not sure though if this is a
>> problem since the TIN used by Summit was created with a very low point
>> density (see below).
>>
>>
>>
>> What DEM did the Summit results use? For Summit, the registration to
>> "truth" for that image is nearly perfect. Was this using the raw camera
>> parameters straight off the platform? Or was the model corrected (using the
>> Summit s/w perhaps)?
>>
>>
>> Summit uses only the PAR file and a TIN. I don't know how Summit uses the
>> information but what I know is that there hasn't been further processing
>> performed on the resulting image after Summit created it.
>>
>> Here's an image showing the topography (black lines) and elevation points
>> (white) used to create the TIN used by Summit:
>> https://www.dropbox.com/s/jivu6xnaws9tnyn/topography.png?dl=0
>>
>> The orange layer in the background is the 1arc DEM and the orthoimage is
>> the one created by Summit. Note the low elevation points density used to
>> create the TIN used by Summit.
>>
>> Thanks for helping me with this. I hope this additional information will
>> be helpful.
>>
>>
>> G
>>
>>
>>
>> Oscar
>>
>> On Monday, May 14, 2018, 8:15:10 AM EDT, Guillaume Drolet <
>> dro...@gm...> wrote:
>>
>>
>> Hi,
>>
>> I try to orthorectify aerial photos from the Ultracam D camera.
>>
>> I have a raw image and a model file (.par) from the DVP software. I copy
>> some of the information from the model file into an Applanix exterior
>> orientation file. The variables I copy from the model file into the
>> Applanix exterior orientation file are:
>> ID, EVENT, EASTING, NORTHING, ORTHOM_HEIGHT, OMEGA, PHI, KAPPA, LAT, LONG
>>
>> I use metadata found in the raw image Tiff tags to write a camera file:
>>
>>
>> NOTE: I set distortion values to zero since I don't have this information.
>>
>> sensor: UCD-SU-1-0034
>> focal_length: 101.400
>> principal_point: -0.180 0.270
>> image_size: 11500.0 7500.0
>> pixel_size: 0.009
>> d1: 1 0
>> d2: 2 0
>> d3: 3 0
>> d4: 4 0
>> d5: 5 0
>> d6: 6 0
>> d7: 7 0
>> d8: 8 0
>> d9: 9 0
>> d10: 10 0
>> d11: 11 0
>> d12: 12 0
>> d13: 13 0
>> d14: 14 0
>> d15: 15 0
>> d16: 16 0
>> d17: 17 0
>> d18: 18 0
>> d19: 19 0
>> d20: 20 0
>> d21: 21 0
>> d22: 22 0
>> d23: 23 0
>> d24: 24 0
>> d25: 25 0
>> d26: 26 0
>> d27: 27 0
>> d28: 28 0
>> d29: 29 0
>> d30: 30 0
>> distortion_units: microns
>>
>>
>> Finally, I use the camera and Applanix files to create a geometry file
>> (.geom) using ossim-applanix2ogeom.
>>
>> When I then run ossim-orthoigen, I don't get results as good as if I use
>> the Summit software to create the orthoimage, feeding it only the raw image
>> and model file. With ossim, the result gets better if I use a 1arc DEM
>> instead of a 100m-DEM but still, it is not as good as when doing it in
>> Summit.
>>
>> I suspect Summit might use more info from the DVP model file which I
>> don't include in my workflow (maybe the rotation matrix?).
>>
>> I would be very grateful if anyone comes up with suggestions that could
>> help achieve results with Ossim that are as good as those obtained using
>> Summit (maybe, for example, by adding more info from the model file into my
>> workflow?
>>
>> I would really like to be able to process my images with ossim: I don't
>> have access to Summit and so I need to rely on other people to do it for me
>> (which isn't always possible).
>>
>> Here are links to images and files that I hope will help you come up with
>> suggestions:
>>
>> raw image: https://www.dropbox.com/s/ s9jdwzsp3a20dpt/JD07137_12007.
>> TIF?dl=0
>> <https://www.dropbox.com/s/s9jdwzsp3a20dpt/JD07137_12007.TIF?dl=0>
>>
>> DVP model file: https://www.dropbox.com/s/
>> eh32r7563a8ll6w/JD07137_12007. par?dl=0
>> <https://www.dropbox.com/s/eh32r7563a8ll6w/JD07137_12007.par?dl=0>
>>
>> geometry file: https://www.dropbox.com/s/ 1qo9kw5fdcsniq1/JD07137_12007.
>> geom?dl=0
>> <https://www.dropbox.com/s/1qo9kw5fdcsniq1/JD07137_12007.geom?dl=0>
>>
>> PNG of orthoimage created with Ossim using a 100m DEM: https://www.dropbox.com/s/
>> zzm40kqo9js4mig/dem_100m.png? dl=0
>> <https://www.dropbox.com/s/zzm40kqo9js4mig/dem_100m.png?dl=0>
>>
>> PNG of orthoimage created with Ossim using a 1arc DEM: https://www.dropbox.com/s/
>> 0x7k4rjbziqb9pp/dem_1arc.png? dl=0
>> <https://www.dropbox.com/s/0x7k4rjbziqb9pp/dem_1arc.png?dl=0>
>>
>> PNG of orthoimage created with Summit: https://www.dropbox.com/s/
>> l8q2qpufq3pkior/summit_TIN. png?dl=0
>> <https://www.dropbox.com/s/l8q2qpufq3pkior/summit_TIN.png?dl=0>
>>
>> Thanks a lot for your help,
>>
>> Guillaume
>>
>>
>>
>>
>>
>>
>>
>>
>> ------------------------------ ------------------------------
>> ------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot______
>> ______________________________ ___________
>> www.ossim.org
>> Ossim-developer mailing list
>> Ossim-developer@lists. sourceforge.net
>> <Oss...@li...>
>> https://lists.sourceforge.net/ lists/listinfo/ossim-developer
>> <https://lists.sourceforge.net/lists/listinfo/ossim-developer>
>>
>>
>>
>

Re: [OSSIM] Orthorectification results: Ossim vs. Summit

[Guillaume D. <dro...@gm...>]

Oscar,

The 1 arcsec DEM has been smoothed using a gaussian filter, hence its
smooth look compared with the 100m DEM. The elevation profiles are for a
diagonal that goes from the lower left corner of the image to its upper
right corner.

Re Summit, as far as I know, Summit is never involved in the
aerotriangulation process. Aerotriangulation is performed as one of the
first steps prior to calculating the affine transformation matrices in DVP
(i.e. the variables in the PAR file).

Also, the TIN used in Summit to orthorectify the image (the one thats
"almost" perfectly lined up with the road network and the Google Earth
image) has much less points than the TIN referenced in the DVP PAR file
(i.e. the .tgl file).

At this point, I suspect that either 1) Summit uses more information from
the DVP PAR file compared with Ossim, e.g. affine transformation matrices,
or 2) Ossim derives transformation matrix coefficients using the Applanix
sensor model, which may not be fully compatible with the UltraCamD camera.

In any case, I will do as you suggested and try to gather more details
about the complete workchain from image acquisition to creation of the DVP
PAR file. I may post more information in this thread if need be.

Thanks a lot for your help,

Guillaume

2018-05-14 16:03 GMT-04:00 Oscar Kramer <osc...@ya...>:

> Guillaume,
>
> RE the DEMS: weird that the 1 arcsec (~30m) is smoother than the 100m DEM.
> I don't know where in that plot your image falls, but there are areas of
> significant bias that could explain the scale difference observed.
>
> RE the Summit code: I have a feeling that your testing is incestuous.
> Perhaps you are using data that was triangulated with Summit using the
> reference image as ground truth, and then using the same model to project
> to the same ground truth. There is no way to explain the perfect alignment
> to the reference image otherwise. It would be a freak of statistical nature
> to have it line up out of the box like that. Check the chain of custody of
> that data set in your organization to see exactly what processing was done
> to it.
>
> Sorry I can't be of more help. I don't have the time to dive into a
> difficult problem like this.
>
> Good luck,
> Oscar
>
> On Monday, May 14, 2018, 1:33:19 PM EDT, Guillaume Drolet <
> dro...@gm...> wrote:
>
>
> Salut Oscar, thanks for your help!
>
> I added more info in your email below. I am not a photogrammetry expert so
> I surely do not understand all the details behind the orthorectification
> process.
>
> G
>
> 2018-05-14 10:19 GMT-04:00 Oscar Kramer <osc...@ya...>:
>
> Salut Guillaume,
>
> Can you give us a little bit of background to help you solve the problem?
> I am not familiar with DVP, PAR nor Summit.
>
>
> DVP is a photogrammetry software used, among other things, to build
> stereoscopic models for pairs of aerial photographs. I don't know if it
> also performs aero-triangulation. It is now discontinued and was replaced,
> in my organization, with Summit (http://www.datem.com/summit-evolution/).
>
> PAR files are model files created by DVP and contain interior and exterior
> orientation parameters. Summit can orthorectify an aerial photograph using
> only the raw image, the corresponding DVP model file (.par), and a
> triangular interpolated network (TIN) file. I don't know if it can also use
> a DEM since I don't work with Summit.
>
> In case you didn't have a look at the PAR file I put in Dropbox, here's
> the content of it:
>
> .GLOBAL
> $DATE 03-03-2008
> $TIME 15:56:18
> $DESCR OPK -> DVP ...
> $TYPE OPK
> $FSCALE00 2.50000000000000e+004
> $FILE00 M:\Dvp\Images\JD07137\opt\rgb\JD07137_12007.tif
> $VERSION 5.5
> .DESCR
> .INTERIOR
> $FCAM00 C:\Projet\G1918\UltraCamD.cam
> $PARAFFINE00 9.00000000000000e-003 1.55854062294791e-018
> -5.15700000000000e+001 8.13151629364128e-020 -9.00000000000000e-003
> 3.34800000000000e+001
> $PARINVAFF00 1.11111111111111e+002 1.92412422586162e-014
> 5.73000000000000e+003 -1.01965198758660e-014 -1.11111111111111e+002
> 3.72000000000000e+003
> $PARM00 1 -5.15700000000000e+001 3.34800000000000e+001
> 0.00000000000000e+000 0.00000000000000e+000 1.42108547152020e-014
> -7.10542735760100e-015 1.42108547152020e-014 -1.42108547152020e-014
> $PARM00 1 -5.15700000000000e+001 3.34800000000000e+001
> 0.00000000000000e+000 0.00000000000000e+000 1.42108547152020e-014
> -7.10542735760100e-015 1.42108547152020e-014 -1.42108547152020e-014
> $PARM00 2 5.19300000000000e+001 3.34800000000000e+001
> 1.15000000000000e+004 0.00000000000000e+000 7.10542735760100e-015
> 7.10542735760100e-015 0.00000000000000e+000 -1.42108547152020e-014
> $PARM00 3 5.19300000000000e+001 -3.40200000000000e+001
> 1.15000000000000e+004 7.50000000000000e+003 -7.10542735760100e-015
> 1.42108547152020e-014 0.00000000000000e+000 0.00000000000000e+000
> $PARM00 4 -5.15700000000000e+001 -3.40200000000000e+001
> 0.00000000000000e+000 7.50000000000000e+003 -7.10542735760100e-015
> 0.00000000000000e+000 1.42108547152020e-014 0.00000000000000e+000
> $DATE 03-03-2008
> $TIME 15:56:18
> $FOC00 1.01400000000000e+002
> .RELATIVE
> .ABSOLUE
> .RELEVEMENT
> .EXTERNE
> .CORRECTION
> .CALIBRAGE
> .TIN
> $TGLNAME M:\Dvp\Mne\JD07137\TGL\JD07137_12007.tgl
> .INDEX
> .LISTMODELES
> .CORRECTION
> .CAMCOR00
> $NAME UltraCamD
> $DATE
> $TYPE 1
> $FOCAL 1.01400000000000e+002
> $PPA 0.00000000000000e+000 0.00000000000000e+000
> $PPS 0.00000000000000e+000 0.00000000000000e+000
> $XP_OFFSET 0.00000000000000e+000
> $YP_OFFSET 0.00000000000000e+000
> $PIXELSIZE 9.00000000000000e+000
> $DISTORTION_TYPE 0
> .EXTERIOR
> $DATE 03-03-2008
> $TIME 15:56:18
> $XYZ00 2.62237954000000e+005 5.40876598900000e+006 2.51442300000000e+003
> $OPK00 -2.24520000000000e-002 -2.24683000000000e+000 8.90675920000000e+001
> .END
>
> I obtained a document explaining part of what this file contains. Here are
> parameter-specific informations:
>
> $PARAFFINE00: contains the affine transformation parameters (a to f) to go
> from image coordinates to photo coordinates. Transformation equations are
> as follows:
> X_photo = a * X_image + b * Y_image + c
> Y_photo = d * X_image + e * Y_image + f
>
> $PARINVAFF0: : contains the inverse affine transformation parameters (A to
> F) to go from photo coordinates to image coordinates. Transformation
> equations are as:
> X_image = A * X_photo + B * Y_photo + C
> Y_image = D * X_photo + E * Y_photo + F
>
> In DVP, origin of image coordinates system is upper left corner.
>
> $XYZ00, $OPK00: exterior orientation parameters. The order of rotations in
> DVP is Omege, Phi and Kappa.
>
> $TGLNAME: this is a DVP generated TIN file. It's a binary file I can't
> read with my tools. However, I have a shapefile version of it (elevation
> points).
>
> It looks like DVP also used information from the file UltraCamD.cam
> (variable $FCAM00 above in the file), maybe to derive the coefficient of
> the affine transformation  ($PARAFFINE00, $PARINVAFF0,  $PARM00 in the file
> above)?
>
> I managed to get the UltraCamD.cam file used in DVP. Here's its content if
> that can help:
>
> SYSTEMAP CAMC120
>
>  101.400
>      -51.750 1       33.750 1
>  -999999.999 0  -999999.999 0
>       51.750 1       33.750 1
>  -999999.999 0  -999999.999 0
>  -999999.999 0  -999999.999 0
>  -999999.999 0  -999999.999 0
>      -51.750 1      -33.750 1
>  -999999.999 0  -999999.999 0
>       51.750 1      -33.750 1
>   1   1
>    -0.180     0.270
>   4 0.0000000e+000 0.0000000e+000 0.0000000e+000 0.0000000e+000
>
>
> I also have a calibration report for the camera but it's a pdf and doesn't
> contain tabular data for distortion values, only plots. Here it is:
> https://www.dropbox.com/s/hm7sr2tpy6c6f2g/Calib-Report_
> 0034_V70_short.pdf?dl=0
>
>
>
>
> Also, what is the ground truth background you are comparing to on the
> mosaics you provided?
>
>
> It's a Google satellite map (OpenLayers QGIS plugin) as well as the road
> network (white lines, 1:20K) produced by my organization (Ministry of
> forests) and used for forest planning, etc.
>
>
>
> There are many things that can be an issue. You pointed out rotation
> matrices. Perhaps the Applanix model if performing the rotations in a
> different order (though I would expect a much bigger discrepancy in that
> case). Also perhaps Summit can read DVP distortion parameters.
>
>
>
> Where would Summit read distortion values from? Could they be the $PARM00
> variables in the PAR file above? As I mentioned, Summit uses only the
> information contained in the PAR file and a TIN to create the orthoimage.
>
>
> RE the 100m v 30 arcsec DEM results. I don't know what to make of that.
> Indeed, the 100 m DEM seems to have a big scale error as if there was a
> significant elevation bias. Did you compare the DEMs to see if there is in
> fact a bias between the two?
>
>
> I created a profile plot that compares the two DEMs (red: 1arc, blue:
> 100m):
>
> https://www.dropbox.com/s/8widxe4lyx657wi/DEM_1arc_vs_100m.png?dl=0
>
> Note how the 1arc DEM has been smoothed. Not sure though if this is a
> problem since the TIN used by Summit was created with a very low point
> density (see below).
>
>
>
> What DEM did the Summit results use? For Summit, the registration to
> "truth" for that image is nearly perfect. Was this using the raw camera
> parameters straight off the platform? Or was the model corrected (using the
> Summit s/w perhaps)?
>
>
> Summit uses only the PAR file and a TIN. I don't know how Summit uses the
> information but what I know is that there hasn't been further processing
> performed on the resulting image after Summit created it.
>
> Here's an image showing the topography (black lines) and elevation points
> (white) used to create the TIN used by Summit:
> https://www.dropbox.com/s/jivu6xnaws9tnyn/topography.png?dl=0
>
> The orange layer in the background is the 1arc DEM and the orthoimage is
> the one created by Summit. Note the low elevation points density used to
> create the TIN used by Summit.
>
> Thanks for helping me with this. I hope this additional information will
> be helpful.
>
>
> G
>
>
>
> Oscar
>
> On Monday, May 14, 2018, 8:15:10 AM EDT, Guillaume Drolet <
> dro...@gm...> wrote:
>
>
> Hi,
>
> I try to orthorectify aerial photos from the Ultracam D camera.
>
> I have a raw image and a model file (.par) from the DVP software. I copy
> some of the information from the model file into an Applanix exterior
> orientation file. The variables I copy from the model file into the
> Applanix exterior orientation file are:
> ID, EVENT, EASTING, NORTHING, ORTHOM_HEIGHT, OMEGA, PHI, KAPPA, LAT, LONG
>
> I use metadata found in the raw image Tiff tags to write a camera file:
>
>
> NOTE: I set distortion values to zero since I don't have this information.
>
> sensor: UCD-SU-1-0034
> focal_length: 101.400
> principal_point: -0.180 0.270
> image_size: 11500.0 7500.0
> pixel_size: 0.009
> d1: 1 0
> d2: 2 0
> d3: 3 0
> d4: 4 0
> d5: 5 0
> d6: 6 0
> d7: 7 0
> d8: 8 0
> d9: 9 0
> d10: 10 0
> d11: 11 0
> d12: 12 0
> d13: 13 0
> d14: 14 0
> d15: 15 0
> d16: 16 0
> d17: 17 0
> d18: 18 0
> d19: 19 0
> d20: 20 0
> d21: 21 0
> d22: 22 0
> d23: 23 0
> d24: 24 0
> d25: 25 0
> d26: 26 0
> d27: 27 0
> d28: 28 0
> d29: 29 0
> d30: 30 0
> distortion_units: microns
>
>
> Finally, I use the camera and Applanix files to create a geometry file
> (.geom) using ossim-applanix2ogeom.
>
> When I then run ossim-orthoigen, I don't get results as good as if I use
> the Summit software to create the orthoimage, feeding it only the raw image
> and model file. With ossim, the result gets better if I use a 1arc DEM
> instead of a 100m-DEM but still, it is not as good as when doing it in
> Summit.
>
> I suspect Summit might use more info from the DVP model file which I don't
> include in my workflow (maybe the rotation matrix?).
>
> I would be very grateful if anyone comes up with suggestions that could
> help achieve results with Ossim that are as good as those obtained using
> Summit (maybe, for example, by adding more info from the model file into my
> workflow?
>
> I would really like to be able to process my images with ossim: I don't
> have access to Summit and so I need to rely on other people to do it for me
> (which isn't always possible).
>
> Here are links to images and files that I hope will help you come up with
> suggestions:
>
> raw image: https://www.dropbox.com/s/ s9jdwzsp3a20dpt/JD07137_12007.
> TIF?dl=0
> <https://www.dropbox.com/s/s9jdwzsp3a20dpt/JD07137_12007.TIF?dl=0>
>
> DVP model file: https://www.dropbox.com/s/ eh32r7563a8ll6w/JD07137_12007.
> par?dl=0
> <https://www.dropbox.com/s/eh32r7563a8ll6w/JD07137_12007.par?dl=0>
>
> geometry file: https://www.dropbox.com/s/ 1qo9kw5fdcsniq1/JD07137_12007.
> geom?dl=0
> <https://www.dropbox.com/s/1qo9kw5fdcsniq1/JD07137_12007.geom?dl=0>
>
> PNG of orthoimage created with Ossim using a 100m DEM: https://www.dropbox.com/s/
> zzm40kqo9js4mig/dem_100m.png? dl=0
> <https://www.dropbox.com/s/zzm40kqo9js4mig/dem_100m.png?dl=0>
>
> PNG of orthoimage created with Ossim using a 1arc DEM: https://www.dropbox.com/s/
> 0x7k4rjbziqb9pp/dem_1arc.png? dl=0
> <https://www.dropbox.com/s/0x7k4rjbziqb9pp/dem_1arc.png?dl=0>
>
> PNG of orthoimage created with Summit: https://www.dropbox.com/s/
> l8q2qpufq3pkior/summit_TIN. png?dl=0
> <https://www.dropbox.com/s/l8q2qpufq3pkior/summit_TIN.png?dl=0>
>
> Thanks a lot for your help,
>
> Guillaume
>
>
>
>
>
>
>
>
> ------------------------------ ------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot______
> ______________________________ ___________
> www.ossim.org
> Ossim-developer mailing list
> Ossim-developer@lists. sourceforge.net
> <Oss...@li...>
> https://lists.sourceforge.net/ lists/listinfo/ossim-developer
> <https://lists.sourceforge.net/lists/listinfo/ossim-developer>
>
>
>

Re: [OSSIM] Orthorectification results: Ossim vs. Summit

[Oscar K. <osc...@ya...>]

 Guillaume,
RE the DEMS: weird that the 1 arcsec (~30m) is smoother than the 100m DEM. I don't know where in that plot your image falls, but there are areas of significant bias that could explain the scale difference observed. 
RE the Summit code: I have a feeling that your testing is incestuous. Perhaps you are using data that was triangulated with Summit using the reference image as ground truth, and then using the same model to project to the same ground truth. There is no way to explain the perfect alignment to the reference image otherwise. It would be a freak of statistical nature to have it line up out of the box like that. Check the chain of custody of that data set in your organization to see exactly what processing was done to it.
Sorry I can't be of more help. I don't have the time to dive into a difficult problem like this.
Good luck,Oscar
    On Monday, May 14, 2018, 1:33:19 PM EDT, Guillaume Drolet <dro...@gm...> wrote:  
 
 Salut Oscar, thanks for your help!

I added more info in your email below. I am not a photogrammetry expert so I surely do not understand all the details behind the orthorectification process.

G

2018-05-14 10:19 GMT-04:00 Oscar Kramer <osc...@ya...>:

 Salut Guillaume,
Can you give us a little bit of background to help you solve the problem? I am not familiar with DVP, PAR nor Summit. 

DVP is a photogrammetry software used, among other things, to build stereoscopic models for pairs of aerial photographs. I don't know if it also performs aero-triangulation.It is now discontinued and was replaced, in my organization, with Summit (http://www.datem.com/summit-evolution/).

PAR files are model files created by DVP and contain interior and exterior orientation parameters. Summit can orthorectify an aerial photograph using only the raw image, the corresponding DVP model file (.par), and a triangular interpolated network (TIN) file. I don't know if it can also use a DEM since I don't work with Summit. 

In case you didn't have a look at the PAR file I put in Dropbox, here's the content of it:

.GLOBAL
$DATE 03-03-2008
$TIME 15:56:18
$DESCR OPK -> DVP ...
$TYPE OPK
$FSCALE00 2.50000000000000e+004
$FILE00 M:\Dvp\Images\JD07137\opt\rgb\JD07137_12007.tif
$VERSION 5.5
.DESCR
.INTERIOR
$FCAM00 C:\Projet\G1918\UltraCamD.cam
$PARAFFINE00 9.00000000000000e-003 1.55854062294791e-018 -5.15700000000000e+001 8.13151629364128e-020 -9.00000000000000e-003 3.34800000000000e+001
$PARINVAFF00 1.11111111111111e+002 1.92412422586162e-014 5.73000000000000e+003 -1.01965198758660e-014 -1.11111111111111e+002 3.72000000000000e+003
$PARM00 1 -5.15700000000000e+001 3.34800000000000e+001 0.00000000000000e+000 0.00000000000000e+000 1.42108547152020e-014 -7.10542735760100e-015 1.42108547152020e-014 -1.42108547152020e-014
$PARM00 1 -5.15700000000000e+001 3.34800000000000e+001 0.00000000000000e+000 0.00000000000000e+000 1.42108547152020e-014 -7.10542735760100e-015 1.42108547152020e-014 -1.42108547152020e-014
$PARM00 2 5.19300000000000e+001 3.34800000000000e+001 1.15000000000000e+004 0.00000000000000e+000 7.10542735760100e-015 7.10542735760100e-015 0.00000000000000e+000 -1.42108547152020e-014
$PARM00 3 5.19300000000000e+001 -3.40200000000000e+001 1.15000000000000e+004 7.50000000000000e+003 -7.10542735760100e-015 1.42108547152020e-014 0.00000000000000e+000 0.00000000000000e+000
$PARM00 4 -5.15700000000000e+001 -3.40200000000000e+001 0.00000000000000e+000 7.50000000000000e+003 -7.10542735760100e-015 0.00000000000000e+000 1.42108547152020e-014 0.00000000000000e+000
$DATE 03-03-2008
$TIME 15:56:18
$FOC00 1.01400000000000e+002
.RELATIVE
.ABSOLUE
.RELEVEMENT
.EXTERNE
.CORRECTION
.CALIBRAGE
.TIN
$TGLNAME M:\Dvp\Mne\JD07137\TGL\JD07137_12007.tgl
.INDEX
.LISTMODELES
.CORRECTION
.CAMCOR00
$NAME UltraCamD
$DATE 
$TYPE 1
$FOCAL 1.01400000000000e+002
$PPA 0.00000000000000e+000 0.00000000000000e+000
$PPS 0.00000000000000e+000 0.00000000000000e+000
$XP_OFFSET 0.00000000000000e+000
$YP_OFFSET 0.00000000000000e+000
$PIXELSIZE 9.00000000000000e+000
$DISTORTION_TYPE 0
.EXTERIOR
$DATE 03-03-2008
$TIME 15:56:18
$XYZ00 2.62237954000000e+005 5.40876598900000e+006 2.51442300000000e+003
$OPK00 -2.24520000000000e-002 -2.24683000000000e+000 8.90675920000000e+001
.END

Iobtained a document explaining part of what this file contains. Here are parameter-specific informations:

$PARAFFINE00: contains the affine transformation parameters (a to f) to go from image coordinates to photo coordinates. Transformation equations are as follows:
X_photo = a * X_image + b * Y_image + c
Y_photo = d * X_image + e * Y_image + f

$PARINVAFF0: : contains the inverse affine transformation parameters (A to F) to go from photo coordinates to image coordinates. Transformation equations are as:
X_image = A * X_photo + B * Y_photo + C
Y_image = D * X_photo + E * Y_photo + F

In DVP, origin of image coordinates system is upper left corner.

$XYZ00, $OPK00: exterior orientation parameters. The order of rotations in DVP is Omege, Phi and Kappa.

$TGLNAME: this is a DVP generated TIN file. It's a binary file I can't read with my tools. However, I have a shapefile version of it (elevation points).

It looks like DVP also used information from the file UltraCamD.cam (variable $FCAM00 above in the file), maybe to derive the coefficient of the affine transformation  ($PARAFFINE00, $PARINVAFF0,  $PARM00 in the file above)?

I managed to get the UltraCamD.cam file used in DVP. Here's its content if that can help:

SYSTEMAP CAMC120

 101.400
     -51.750 1       33.750 1
 -999999.999 0  -999999.999 0
      51.750 1       33.750 1
 -999999.999 0  -999999.999 0
 -999999.999 0  -999999.999 0
 -999999.999 0  -999999.999 0
     -51.750 1      -33.750 1
 -999999.999 0  -999999.999 0
      51.750 1      -33.750 1
  1   1
   -0.180     0.270
  4 0.0000000e+000 0.0000000e+000 0.0000000e+000 0.0000000e+000


I also have a calibration report for the camera but it's a pdf and doesn't contain tabular data for distortion values, only plots. Here it is:
https://www.dropbox.com/s/hm7sr2tpy6c6f2g/Calib-Report_0034_V70_short.pdf?dl=0


 
Also, what is the ground truth background you are comparing to on the mosaics you provided?

It's a Google satellite map (OpenLayers QGIS plugin) as well as the road network (white lines, 1:20K) produced by my organization (Ministry of forests) and used for forest planning, etc.




There are many things that can be an issue. You pointed out rotation matrices. Perhaps the Applanix model if performing the rotations in a different order (though I would expect a much bigger discrepancy in that case). Also perhaps Summit can read DVP distortion parameters.


Where would Summit read distortion values from? Could they be the $PARM00  variables in the PAR file above? As I mentioned, Summit uses only the information contained in the PAR file and a TIN to create the orthoimage.



RE the 100m v 30 arcsec DEM results. I don't know what to make of that. Indeed, the 100 m DEM seems to have a big scale error as if there was a significant elevation bias. Did you compare the DEMs to see if there is in fact a bias between the two? 

I created a profile plot that compares the two DEMs (red: 1arc, blue: 100m):

https://www.dropbox.com/s/8widxe4lyx657wi/DEM_1arc_vs_100m.png?dl=0

Note how the 1arc DEM has been smoothed. Not sure though if this is a problem since the TIN used by Summit was created with a very low point density (see below).

 
What DEM did the Summit results use? For Summit, the registration to "truth" for that image is nearly perfect. Was this using the raw camera parameters straight off the platform? Or was the model corrected (using the Summit s/w perhaps)?

Summit uses only the PAR file and a TIN. I don't know how Summit uses the information but what I know is that there hasn't been further processing performed on the resulting image after Summit created it.  

Here's an image showing the topography (black lines) and elevation points (white) used to create the TIN used by Summit:
https://www.dropbox.com/s/jivu6xnaws9tnyn/topography.png?dl=0

The orange layer in the background is the 1arc DEM and the orthoimage is the one created by Summit. Note the low elevation points density used to create the TIN used by Summit.

Thanks for helping me with this. I hope this additional information will be helpful.

G




Oscar
    On Monday, May 14, 2018, 8:15:10 AM EDT, Guillaume Drolet <dro...@gm...> wrote:  
 
 Hi,

I try to orthorectify aerial photos from the Ultracam D camera.

I have a raw image and a model file (.par) from the DVP software. I copy some of the information from the model file into an Applanix exterior orientation file. The variables I copy from the model file into the Applanix exterior orientation file are:
ID, EVENT, EASTING, NORTHING, ORTHOM_HEIGHT, OMEGA, PHI, KAPPA, LAT, LONG

I use metadata found in the raw image Tiff tags to write a camera file:


NOTE: I set distortion values to zero since I don't have this information.

sensor: UCD-SU-1-0034
focal_length: 101.400
principal_point: -0.180 0.270
image_size: 11500.0 7500.0
pixel_size: 0.009
d1: 1 0
d2: 2 0
d3: 3 0
d4: 4 0
d5: 5 0
d6: 6 0
d7: 7 0
d8: 8 0
d9: 9 0
d10: 10 0
d11: 11 0
d12: 12 0
d13: 13 0
d14: 14 0
d15: 15 0
d16: 16 0
d17: 17 0
d18: 18 0
d19: 19 0
d20: 20 0
d21: 21 0
d22: 22 0
d23: 23 0
d24: 24 0
d25: 25 0
d26: 26 0
d27: 27 0
d28: 28 0
d29: 29 0
d30: 30 0
distortion_units: microns


Finally, I use the camera and Applanix files to create a geometry file (.geom) using ossim-applanix2ogeom.

When I then run ossim-orthoigen, I don't get results as good as if I use theSummitsoftware to create the orthoimage, feeding it only the raw image and model file. With ossim, the result gets better if I use a 1arc DEM instead of a 100m-DEM but still, it is not as good as when doing it in Summit.

I suspect Summit might use more info from the DVP model file which I don't include in my workflow (maybe the rotation matrix?). 

I would be very grateful if anyone comes up with suggestions that could help achieve results with Ossim that are as good as those obtained using Summit (maybe, for example, by adding more info from the model file into my workflow? 

I would really like to be able to process my images with ossim: I don't have access to Summit and so I need to rely on other people to do it for me (which isn't always possible).

Here are links to images and files that I hope will help you come up with suggestions:

raw image: https://www.dropbox.com/s/ s9jdwzsp3a20dpt/JD07137_12007. TIF?dl=0

DVP model file: https://www.dropbox.com/s/ eh32r7563a8ll6w/JD07137_12007. par?dl=0

geometry file: https://www.dropbox.com/s/ 1qo9kw5fdcsniq1/JD07137_12007. geom?dl=0

PNG of orthoimage created with Ossim using a 100m DEM: https://www.dropbox.com/s/ zzm40kqo9js4mig/dem_100m.png? dl=0

PNG of orthoimage created with Ossim using a 1arc DEM: https://www.dropbox.com/s/ 0x7k4rjbziqb9pp/dem_1arc.png? dl=0

PNG of orthoimage created with Summit: https://www.dropbox.com/s/ l8q2qpufq3pkior/summit_TIN. png?dl=0

Thanks a lot for your help,

Guillaume








------------------------------ ------------------------------ ------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot______ ______________________________ ___________
www.ossim.org
Ossim-developer mailing list
Ossim-developer@lists. sourceforge.net
https://lists.sourceforge.net/ lists/listinfo/ossim-developer
  

  

Re: [OSSIM] Orthorectification results: Ossim vs. Summit

[Guillaume D. <dro...@gm...>]

Salut Oscar, thanks for your help!

I added more info in your email below. I am not a photogrammetry expert so
I surely do not understand all the details behind the orthorectification
process.

G

2018-05-14 10:19 GMT-04:00 Oscar Kramer <osc...@ya...>:

> Salut Guillaume,
>
> Can you give us a little bit of background to help you solve the problem?
> I am not familiar with DVP, PAR nor Summit.
>

DVP is a photogrammetry software used, among other things, to build
stereoscopic models for pairs of aerial photographs. I don't know if it
also performs aero-triangulation. It is now discontinued and was replaced,
in my organization, with Summit (http://www.datem.com/summit-evolution/).

PAR files are model files created by DVP and contain interior and exterior
orientation parameters. Summit can orthorectify an aerial photograph using
only the raw image, the corresponding DVP model file (.par), and a
triangular interpolated network (TIN) file. I don't know if it can also use
a DEM since I don't work with Summit.

In case you didn't have a look at the PAR file I put in Dropbox, here's the
content of it:

.GLOBAL
$DATE 03-03-2008
$TIME 15:56:18
$DESCR OPK -> DVP ...
$TYPE OPK
$FSCALE00 2.50000000000000e+004
$FILE00 M:\Dvp\Images\JD07137\opt\rgb\JD07137_12007.tif
$VERSION 5.5
.DESCR
.INTERIOR
$FCAM00 C:\Projet\G1918\UltraCamD.cam
$PARAFFINE00 9.00000000000000e-003 1.55854062294791e-018
-5.15700000000000e+001 8.13151629364128e-020 -9.00000000000000e-003
3.34800000000000e+001
$PARINVAFF00 1.11111111111111e+002 1.92412422586162e-014
5.73000000000000e+003 -1.01965198758660e-014 -1.11111111111111e+002
3.72000000000000e+003
$PARM00 1 -5.15700000000000e+001 3.34800000000000e+001
0.00000000000000e+000 0.00000000000000e+000 1.42108547152020e-014
-7.10542735760100e-015 1.42108547152020e-014 -1.42108547152020e-014
$PARM00 1 -5.15700000000000e+001 3.34800000000000e+001
0.00000000000000e+000 0.00000000000000e+000 1.42108547152020e-014
-7.10542735760100e-015 1.42108547152020e-014 -1.42108547152020e-014
$PARM00 2 5.19300000000000e+001 3.34800000000000e+001 1.15000000000000e+004
0.00000000000000e+000 7.10542735760100e-015 7.10542735760100e-015
0.00000000000000e+000 -1.42108547152020e-014
$PARM00 3 5.19300000000000e+001 -3.40200000000000e+001
1.15000000000000e+004 7.50000000000000e+003 -7.10542735760100e-015
1.42108547152020e-014 0.00000000000000e+000 0.00000000000000e+000
$PARM00 4 -5.15700000000000e+001 -3.40200000000000e+001
0.00000000000000e+000 7.50000000000000e+003 -7.10542735760100e-015
0.00000000000000e+000 1.42108547152020e-014 0.00000000000000e+000
$DATE 03-03-2008
$TIME 15:56:18
$FOC00 1.01400000000000e+002
.RELATIVE
.ABSOLUE
.RELEVEMENT
.EXTERNE
.CORRECTION
.CALIBRAGE
.TIN
$TGLNAME M:\Dvp\Mne\JD07137\TGL\JD07137_12007.tgl
.INDEX
.LISTMODELES
.CORRECTION
.CAMCOR00
$NAME UltraCamD
$DATE
$TYPE 1
$FOCAL 1.01400000000000e+002
$PPA 0.00000000000000e+000 0.00000000000000e+000
$PPS 0.00000000000000e+000 0.00000000000000e+000
$XP_OFFSET 0.00000000000000e+000
$YP_OFFSET 0.00000000000000e+000
$PIXELSIZE 9.00000000000000e+000
$DISTORTION_TYPE 0
.EXTERIOR
$DATE 03-03-2008
$TIME 15:56:18
$XYZ00 2.62237954000000e+005 5.40876598900000e+006 2.51442300000000e+003
$OPK00 -2.24520000000000e-002 -2.24683000000000e+000 8.90675920000000e+001
.END

I obtained a document explaining part of what this file contains. Here are
parameter-specific informations:

$PARAFFINE00: contains the affine transformation parameters (a to f) to go
from image coordinates to photo coordinates. Transformation equations are
as follows:
X_photo = a * X_image + b * Y_image + c
Y_photo = d * X_image + e * Y_image + f

$PARINVAFF0: : contains the inverse affine transformation parameters (A to
F) to go from photo coordinates to image coordinates. Transformation
equations are as:
X_image = A * X_photo + B * Y_photo + C
Y_image = D * X_photo + E * Y_photo + F

In DVP, origin of image coordinates system is upper left corner.

$XYZ00, $OPK00: exterior orientation parameters. The order of rotations in
DVP is Omege, Phi and Kappa.

$TGLNAME: this is a DVP generated TIN file. It's a binary file I can't read
with my tools. However, I have a shapefile version of it (elevation points).

It looks like DVP also used information from the file UltraCamD.cam
(variable $FCAM00 above in the file), maybe to derive the coefficient of
the affine transformation  ($PARAFFINE00, $PARINVAFF0,  $PARM00 in the file
above)?

I managed to get the UltraCamD.cam file used in DVP. Here's its content if
that can help:

SYSTEMAP CAMC120

 101.400
     -51.750 1       33.750 1
 -999999.999 0  -999999.999 0
      51.750 1       33.750 1
 -999999.999 0  -999999.999 0
 -999999.999 0  -999999.999 0
 -999999.999 0  -999999.999 0
     -51.750 1      -33.750 1
 -999999.999 0  -999999.999 0
      51.750 1      -33.750 1
  1   1
   -0.180     0.270
  4 0.0000000e+000 0.0000000e+000 0.0000000e+000 0.0000000e+000


I also have a calibration report for the camera but it's a pdf and doesn't
contain tabular data for distortion values, only plots. Here it is:
https://www.dropbox.com/s/hm7sr2tpy6c6f2g/Calib-Report_0034_V70_short.pdf?dl=0




> Also, what is the ground truth background you are comparing to on the
> mosaics you provided?
>

It's a Google satellite map (OpenLayers QGIS plugin) as well as the road
network (white lines, 1:20K) produced by my organization (Ministry of
forests) and used for forest planning, etc.



> There are many things that can be an issue. You pointed out rotation
> matrices. Perhaps the Applanix model if performing the rotations in a
> different order (though I would expect a much bigger discrepancy in that
> case). Also perhaps Summit can read DVP distortion parameters.
>


Where would Summit read distortion values from? Could they be the $PARM00
variables in the PAR file above? As I mentioned, Summit uses only the
information contained in the PAR file and a TIN to create the orthoimage.


> RE the 100m v 30 arcsec DEM results. I don't know what to make of that.
> Indeed, the 100 m DEM seems to have a big scale error as if there was a
> significant elevation bias. Did you compare the DEMs to see if there is in
> fact a bias between the two?
>

I created a profile plot that compares the two DEMs (red: 1arc, blue: 100m):

https://www.dropbox.com/s/8widxe4lyx657wi/DEM_1arc_vs_100m.png?dl=0

Note how the 1arc DEM has been smoothed. Not sure though if this is a
problem since the TIN used by Summit was created with a very low point
density (see below).



> What DEM did the Summit results use? For Summit, the registration to
> "truth" for that image is nearly perfect. Was this using the raw camera
> parameters straight off the platform? Or was the model corrected (using the
> Summit s/w perhaps)?
>

Summit uses only the PAR file and a TIN. I don't know how Summit uses the
information but what I know is that there hasn't been further processing
performed on the resulting image after Summit created it.

Here's an image showing the topography (black lines) and elevation points
(white) used to create the TIN used by Summit:
https://www.dropbox.com/s/jivu6xnaws9tnyn/topography.png?dl=0

The orange layer in the background is the 1arc DEM and the orthoimage is
the one created by Summit. Note the low elevation points density used to
create the TIN used by Summit.

Thanks for helping me with this. I hope this additional information will be
helpful.

G



> Oscar
>
> On Monday, May 14, 2018, 8:15:10 AM EDT, Guillaume Drolet <
> dro...@gm...> wrote:
>
>
> Hi,
>
> I try to orthorectify aerial photos from the Ultracam D camera.
>
> I have a raw image and a model file (.par) from the DVP software. I copy
> some of the information from the model file into an Applanix exterior
> orientation file. The variables I copy from the model file into the
> Applanix exterior orientation file are:
> ID, EVENT, EASTING, NORTHING, ORTHOM_HEIGHT, OMEGA, PHI, KAPPA, LAT, LONG
>
> I use metadata found in the raw image Tiff tags to write a camera file:
>
>
> NOTE: I set distortion values to zero since I don't have this information.
>
> sensor: UCD-SU-1-0034
> focal_length: 101.400
> principal_point: -0.180 0.270
> image_size: 11500.0 7500.0
> pixel_size: 0.009
> d1: 1 0
> d2: 2 0
> d3: 3 0
> d4: 4 0
> d5: 5 0
> d6: 6 0
> d7: 7 0
> d8: 8 0
> d9: 9 0
> d10: 10 0
> d11: 11 0
> d12: 12 0
> d13: 13 0
> d14: 14 0
> d15: 15 0
> d16: 16 0
> d17: 17 0
> d18: 18 0
> d19: 19 0
> d20: 20 0
> d21: 21 0
> d22: 22 0
> d23: 23 0
> d24: 24 0
> d25: 25 0
> d26: 26 0
> d27: 27 0
> d28: 28 0
> d29: 29 0
> d30: 30 0
> distortion_units: microns
>
>
> Finally, I use the camera and Applanix files to create a geometry file
> (.geom) using ossim-applanix2ogeom.
>
> When I then run ossim-orthoigen, I don't get results as good as if I use
> the Summit software to create the orthoimage, feeding it only the raw image
> and model file. With ossim, the result gets better if I use a 1arc DEM
> instead of a 100m-DEM but still, it is not as good as when doing it in
> Summit.
>
> I suspect Summit might use more info from the DVP model file which I don't
> include in my workflow (maybe the rotation matrix?).
>
> I would be very grateful if anyone comes up with suggestions that could
> help achieve results with Ossim that are as good as those obtained using
> Summit (maybe, for example, by adding more info from the model file into my
> workflow?
>
> I would really like to be able to process my images with ossim: I don't
> have access to Summit and so I need to rely on other people to do it for me
> (which isn't always possible).
>
> Here are links to images and files that I hope will help you come up with
> suggestions:
>
> raw image: https://www.dropbox.com/s/s9jdwzsp3a20dpt/JD07137_12007.
> TIF?dl=0
>
> DVP model file: https://www.dropbox.com/s/eh32r7563a8ll6w/JD07137_12007.
> par?dl=0
>
> geometry file: https://www.dropbox.com/s/1qo9kw5fdcsniq1/JD07137_12007.
> geom?dl=0
>
> PNG of orthoimage created with Ossim using a 100m DEM:
> https://www.dropbox.com/s/zzm40kqo9js4mig/dem_100m.png?dl=0
>
> PNG of orthoimage created with Ossim using a 1arc DEM:
> https://www.dropbox.com/s/0x7k4rjbziqb9pp/dem_1arc.png?dl=0
>
> PNG of orthoimage created with Summit: https://www.dropbox.com/s/
> l8q2qpufq3pkior/summit_TIN.png?dl=0
>
> Thanks a lot for your help,
>
> Guillaume
>
>
>
>
>
>
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot______
> _________________________________________
> www.ossim.org
> Ossim-developer mailing list
> Oss...@li...
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>

Re: [OSSIM] Orthorectification results: Ossim vs. Summit

[Oscar K. <osc...@ya...>]

 Salut Guillaume,
Can you give us a little bit of background to help you solve the problem? I am not familiar with DVP, PAR nor Summit. Also, what is the ground truth background you are comparing to on the mosaics you provided?
There are many things that can be an issue. You pointed out rotation matrices. Perhaps the Applanix model if performing the rotations in a different order (though I would expect a much bigger discrepancy in that case). Also perhaps Summit can read DVP distortion parameters.
RE the 100m v 30 arcsec DEM results. I don't know what to make of that. Indeed, the 100 m DEM seems to have a big scale error as if there was a significant elevation bias. Did you compare the DEMs to see if there is in fact a bias between the two? What DEM did the Summit results use? For Summit, the registration to "truth" for that image is nearly perfect. Was this using the raw camera parameters straight off the platform? Or was the model corrected (using the Summit s/w perhaps)?
Oscar
    On Monday, May 14, 2018, 8:15:10 AM EDT, Guillaume Drolet <dro...@gm...> wrote:  
 
 Hi,

I try to orthorectify aerial photos from the Ultracam D camera.

I have a raw image and a model file (.par) from the DVP software. I copy some of the information from the model file into an Applanix exterior orientation file. The variables I copy from the model file into the Applanix exterior orientation file are:
ID, EVENT, EASTING, NORTHING, ORTHOM_HEIGHT, OMEGA, PHI, KAPPA, LAT, LONG

I use metadata found in the raw image Tiff tags to write a camera file:


NOTE: I set distortion values to zero since I don't have this information.

sensor: UCD-SU-1-0034
focal_length: 101.400
principal_point: -0.180 0.270
image_size: 11500.0 7500.0
pixel_size: 0.009
d1: 1 0
d2: 2 0
d3: 3 0
d4: 4 0
d5: 5 0
d6: 6 0
d7: 7 0
d8: 8 0
d9: 9 0
d10: 10 0
d11: 11 0
d12: 12 0
d13: 13 0
d14: 14 0
d15: 15 0
d16: 16 0
d17: 17 0
d18: 18 0
d19: 19 0
d20: 20 0
d21: 21 0
d22: 22 0
d23: 23 0
d24: 24 0
d25: 25 0
d26: 26 0
d27: 27 0
d28: 28 0
d29: 29 0
d30: 30 0
distortion_units: microns


Finally, I use the camera and Applanix files to create a geometry file (.geom) using ossim-applanix2ogeom.

When I then run ossim-orthoigen, I don't get results as good as if I use theSummitsoftware to create the orthoimage, feeding it only the raw image and model file. With ossim, the result gets better if I use a 1arc DEM instead of a 100m-DEM but still, it is not as good as when doing it in Summit.

I suspect Summit might use more info from the DVP model file which I don't include in my workflow (maybe the rotation matrix?). 

I would be very grateful if anyone comes up with suggestions that could help achieve results with Ossim that are as good as those obtained using Summit (maybe, for example, by adding more info from the model file into my workflow? 

I would really like to be able to process my images with ossim: I don't have access to Summit and so I need to rely on other people to do it for me (which isn't always possible).

Here are links to images and files that I hope will help you come up with suggestions:

raw image: https://www.dropbox.com/s/s9jdwzsp3a20dpt/JD07137_12007.TIF?dl=0

DVP model file: https://www.dropbox.com/s/eh32r7563a8ll6w/JD07137_12007.par?dl=0

geometry file: https://www.dropbox.com/s/1qo9kw5fdcsniq1/JD07137_12007.geom?dl=0

PNG of orthoimage created with Ossim using a 100m DEM: https://www.dropbox.com/s/zzm40kqo9js4mig/dem_100m.png?dl=0

PNG of orthoimage created with Ossim using a 1arc DEM: https://www.dropbox.com/s/0x7k4rjbziqb9pp/dem_1arc.png?dl=0

PNG of orthoimage created with Summit: https://www.dropbox.com/s/l8q2qpufq3pkior/summit_TIN.png?dl=0

Thanks a lot for your help,

Guillaume








------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot_______________________________________________
www.ossim.org
Ossim-developer mailing list
Oss...@li...
https://lists.sourceforge.net/lists/listinfo/ossim-developer
  

[OSSIM] Orthorectification results: Ossim vs. Summit

[Guillaume D. <dro...@gm...>]

Hi,

I try to orthorectify aerial photos from the Ultracam D camera.

I have a raw image and a model file (.par) from the DVP software. I copy
some of the information from the model file into an Applanix exterior
orientation file. The variables I copy from the model file into the
Applanix exterior orientation file are:
ID, EVENT, EASTING, NORTHING, ORTHOM_HEIGHT, OMEGA, PHI, KAPPA, LAT, LONG

I use metadata found in the raw image Tiff tags to write a camera file:


NOTE: I set distortion values to zero since I don't have this information.

sensor: UCD-SU-1-0034
focal_length: 101.400
principal_point: -0.180 0.270
image_size: 11500.0 7500.0
pixel_size: 0.009
d1: 1 0
d2: 2 0
d3: 3 0
d4: 4 0
d5: 5 0
d6: 6 0
d7: 7 0
d8: 8 0
d9: 9 0
d10: 10 0
d11: 11 0
d12: 12 0
d13: 13 0
d14: 14 0
d15: 15 0
d16: 16 0
d17: 17 0
d18: 18 0
d19: 19 0
d20: 20 0
d21: 21 0
d22: 22 0
d23: 23 0
d24: 24 0
d25: 25 0
d26: 26 0
d27: 27 0
d28: 28 0
d29: 29 0
d30: 30 0
distortion_units: microns


Finally, I use the camera and Applanix files to create a geometry file
(.geom) using ossim-applanix2ogeom.

When I then run ossim-orthoigen, I don't get results as good as if I use
the Summit software to create the orthoimage, feeding it only the raw image
and model file. With ossim, the result gets better if I use a 1arc DEM
instead of a 100m-DEM but still, it is not as good as when doing it in
Summit.

I suspect Summit might use more info from the DVP model file which I don't
include in my workflow (maybe the rotation matrix?).

I would be very grateful if anyone comes up with suggestions that could
help achieve results with Ossim that are as good as those obtained using
Summit (maybe, for example, by adding more info from the model file into my
workflow?

I would really like to be able to process my images with ossim: I don't
have access to Summit and so I need to rely on other people to do it for me
(which isn't always possible).

Here are links to images and files that I hope will help you come up with
suggestions:

raw image: https://www.dropbox.com/s/s9jdwzsp3a20dpt/JD07137_12007.TIF?dl=0

DVP model file:
https://www.dropbox.com/s/eh32r7563a8ll6w/JD07137_12007.par?dl=0

geometry file:
https://www.dropbox.com/s/1qo9kw5fdcsniq1/JD07137_12007.geom?dl=0

PNG of orthoimage created with Ossim using a 100m DEM:
https://www.dropbox.com/s/zzm40kqo9js4mig/dem_100m.png?dl=0

PNG of orthoimage created with Ossim using a 1arc DEM:
https://www.dropbox.com/s/0x7k4rjbziqb9pp/dem_1arc.png?dl=0

PNG of orthoimage created with Summit:
https://www.dropbox.com/s/l8q2qpufq3pkior/summit_TIN.png?dl=0

Thanks a lot for your help,

Guillaume

Re: [OSSIM] Fwd: orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read Error! t

[Guillaume D. <dro...@gm...>]

ok, thanks for making the tests with my image.

so, does this mean the libtiff library that was linked to ossim during
compilation didn't have compression support and there's nothing I can do
about it?

do you know if there is somewhere windows binaries with compression support
enabled that I could use instead?



On Sat, May 12, 2018, 12:22 David Burken, <db...@gm...> wrote:

> ---------- Forwarded message ----------
> From: David Burken <db...@gm...>
> Date: Sat, 12 May 2018 12:18:23 -0400
> Subject: Re: [OSSIM] orthoigen error:
> ossimTiffTileSource::loadFromRgbaTile Read Error! t
> To: Guillaume Drolet <dro...@gm...>
>
> All, looks good... Looks good in geocell, cmm, preproc and chipper
> sequenced through fine. See attached....
>
> $ ossim-preproc -o --ch JD07137_12007.TIF
> Processing file: /home/dburken/Downloads/JD07137_12007.TIF
> Creating overviews with histogram for file:
> /home/dburken/Downloads/JD07137_12007.TIF
> creating r1...
> 100%
> ...
> creating r11...
> 100%
> Finished...
>
>
> [dburken@pluto Downloads]$ ossim-cmm JD07137_12007.TIF
> Processing image file:  JD07137_12007.TIF
> Computing min/max for entry number:  0
> 100%
> Finished...
> wrote file:  JD07137_12007.omd
>
> $ ossim-chipper --op chip -t 256 JD07137_12007.TIF t1.png
> 100%
> elapsed time in seconds: 0.972
>
>
> On 5/12/18, Guillaume Drolet <dro...@gm...> wrote:
> > Did you try, for example, running ossim-cmm on the image?
> >
> > ossim-info executes fine but I get the error when running ossim-cmm or
> > ossim-orthoigen.
> >
> > On Sat, May 12, 2018, 11:28 David Burken, <db...@gm...> wrote:
> >
> >> Hi Guilaume,
> >>
> >> I can bring up the file fine. No projection but I can view. No
> geographic
> >> tags in there. So that points to your libtiff maybe?
> >>
> >> Hope that helps,
> >> Dave
> >>
> >>
> >> On Fri, May 11, 2018 at 12:38 PM, Guillaume Drolet <
> >> dro...@gm...> wrote:
> >>
> >>>
> >>> Hi,
> >>>
> >>> I fixed my problem by creating a new version of my input tif file using
> >>> gdal_translate.
> >>>
> >>> I don't know what was causing the error but I suspect OSSIM doesn't
> like
> >>> compressed tiff files: input file had COMPRESSION=YCbCr JPEG prior to
> >>> running gdal_translate (no compression in the resulting file).
> >>>
> >>> Can anyone confirm this?
> >>>
> >>>
> >>>
> >>>
> >>> *[OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read
> >>> Error! t*
> >>> From: Guillaume Drolet <droletguillaume@gm...> - 2018-05-10 15:40:48
> >>>
> >>>  Hello Everyone,
> >>>
> >>> I few years ago I used ossim-orthoigen for orthorectification of aerial
> >>> photographs and I got great help from your list
> >>> (http://osgeo-org.1560.x6.nabble.com/Orthorectification-of-
> >>> UltraCAM-D-images-td5063637.html)
> >>>
> >>> I recently re-installed OSSIM version 1.8.19 from Windows binaries I
> >>> found
> >>> on the web (ossim_x64_1_8_19-20150428-2.exe) to orthorectify new images
> >>> from the same camera. My elevation setup is ok and I tested
> >>> ossim-orthoigen
> >>> on the same image (.TIF) and geometry file I used 4 years ago. But now
> I
> >>> get this error message (excerpt from tracing ossimTiffTileSource):
> >>>
> >>> ossimSource::print:
> >>> theEnableFlag:       1
> >>> theInitializedFlag:  0
> >>> ossimErrorStatusInterface::print
> >>> theErrorStatus:         0
> >>> theErrorStatus string:  OSSIM_OK
> >>> ossimTiffTileSource::allocateBuffer DEBUG:
> >>> buffer_size:  88176
> >>> ossimTiffTileSource::allocateBuffer DEBUG:
> >>> buffer_size:  262144
> >>> ossimTiffTileSource::loadFromRgbaTile Read Error!
> >>> Returning error...
> >>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> >>> buffer. Return status = false...
> >>> ossimTiffTileSource::loadFromRgbaTile Read Error!
> >>> Returning error...
> >>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> >>> buffer. Return status = false...
> >>> ossimTiffTileSource::loadFromRgbaTile Read Error!
> >>> Returning error...
> >>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> >>> buffer. Return status = false...
> >>> ossimTiffTileSource::loadFromRgbaTile Read Error!
> >>> Returning error...
> >>>
> >>> I skimmed the source code but don't have a setup to perform debugging.
> >>> Do
> >>> you have any ideas what could be the source of this error?
> >>>
> >>> If this can be helpful, here are the first few lines of the log file
> >>> (--ossim-logfile):
> >>>
> >>> ossimTiffTileSource::open Entered...
> >>> File:  JD07137_12007.TIF
> >>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> >>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> >>> theMinSampleValue:  -1.#QNAN
> >>> theMaxSampleValue:  -1.#QNAN
> >>> ossimTiffTileSource::open Debug:image_file:
> >>> JD07137_12007.TIF
> >>> samples_per_pixel:           3
> >>> bits_per_sample:             8
> >>> sample_format_unit:          0
> >>> min_sample_value:            1
> >>> max_sample_value:            255
> >>> null_sample_value:           0
> >>> theNumberOfDirectories:      7
> >>> r0_is_full_res:              1
> >>> directory[0]
> >>> image width:     11500
> >>> image_length:    7500
> >>> read method:     READ_RGBA_U8_TILE
> >>> planar:          1
> >>> photometric:     6
> >>> tile_width:      256
> >>> tile_length:     256
> >>>
> >>> directory[1]
> >>> image width:     5750
> >>> image_length:    3750
> >>> read method:     READ_RGBA_U8_TILE
> >>> planar:          1
> >>> photometric:     6
> >>> tile_width:      256
> >>> tile_length:     256
> >>>
> >>> directory[2]
> >>> image width:     2875
> >>> image_length:    1875
> >>> read method:     READ_RGBA_U8_TILE
> >>> planar:          1
> >>> photometric:     6
> >>> tile_width:      256
> >>> tile_length:     256
> >>>
> >>> ...
> >>>
> >>> ossimSource::print:
> >>> theEnableFlag:       1
> >>> theInitializedFlag:  0
> >>> ossimErrorStatusInterface::print
> >>> theErrorStatus:         0
> >>> theErrorStatus string:  OSSIM_OK
> >>> ossimTiffTileSource::open Entered...
> >>> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> >>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> >>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> >>> theMinSampleValue:  -1.#QNAN
> >>> theMaxSampleValue:  -1.#QNAN
> >>> ossimTiffTileSource::open Debug:image_file:
> >>> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> >>> samples_per_pixel:           1
> >>> bits_per_sample:             32
> >>> sample_format_unit:          3
> >>> min_sample_value:            -8.38861e+006
> >>> max_sample_value:            8.38861e+006
> >>> null_sample_value:           -99999
> >>> theNumberOfDirectories:      1
> >>> r0_is_full_res:              1
> >>> directory[0]
> >>> image width:     11022
> >>> image_length:    9833
> >>> read method:     READ_SCAN_LINE
> >>> planar:          1
> >>> photometric:     1
> >>> rows per strip:  1
> >>> tile_width:      256
> >>> tile_length:     64
> >>>
> >>> ossimSource::print:
> >>> theEnableFlag:       1
> >>> theInitializedFlag:  0
> >>> ossimErrorStatusInterface::print
> >>> theErrorStatus:         0
> >>> theErrorStatus string:  OSSIM_OK
> >>> ossimTiffTileSource::open Entered...
> >>> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> >>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> >>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> >>> theMinSampleValue:  -1.#QNAN
> >>> theMaxSampleValue:  -1.#QNAN
> >>> ossimTiffTileSource::open Debug:image_file:
> >>> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> >>> samples_per_pixel:           1
> >>> bits_per_sample:             32
> >>> sample_format_unit:          3
> >>> min_sample_value:            -8.38861e+006
> >>> max_sample_value:            8.38861e+006
> >>> null_sample_value:           -99999
> >>> theNumberOfDirectories:      1
> >>> r0_is_full_res:              1
> >>> directory[0]
> >>> image width:     11022
> >>> image_length:    9833
> >>> read method:     READ_SCAN_LINE
> >>> planar:          1
> >>> photometric:     1
> >>> rows per strip:  1
> >>> tile_width:      256
> >>> tile_length:     64
> >>>
> >>> ...
> >>>
> >>>
> >>> I will appreciate any advice on the possible source(s) of this error.
> >>>
> >>> Best regards,
> >>> Guillaume
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> ------------------------------------------------------------------------------
> >>> Check out the vibrant tech community on one of the world's most
> >>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> >>> _______________________________________________
> >>> www.ossim.org
> >>> Ossim-developer mailing list
> >>> Oss...@li...
> >>> https://lists.sourceforge.net/lists/listinfo/ossim-developer
> >>>
> >>>
> >>
> >
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> Oss...@li...
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>

[OSSIM] Fwd: orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read Error! t

[David B. <db...@gm...>]

---------- Forwarded message ----------
From: David Burken <db...@gm...>
Date: Sat, 12 May 2018 12:18:23 -0400
Subject: Re: [OSSIM] orthoigen error:
ossimTiffTileSource::loadFromRgbaTile Read Error! t
To: Guillaume Drolet <dro...@gm...>

All, looks good... Looks good in geocell, cmm, preproc and chipper
sequenced through fine. See attached....

$ ossim-preproc -o --ch JD07137_12007.TIF
Processing file: /home/dburken/Downloads/JD07137_12007.TIF
Creating overviews with histogram for file:
/home/dburken/Downloads/JD07137_12007.TIF
creating r1...
100%
...
creating r11...
100%
Finished...


[dburken@pluto Downloads]$ ossim-cmm JD07137_12007.TIF
Processing image file:  JD07137_12007.TIF
Computing min/max for entry number:  0
100%
Finished...
wrote file:  JD07137_12007.omd

$ ossim-chipper --op chip -t 256 JD07137_12007.TIF t1.png
100%
elapsed time in seconds: 0.972


On 5/12/18, Guillaume Drolet <dro...@gm...> wrote:
> Did you try, for example, running ossim-cmm on the image?
>
> ossim-info executes fine but I get the error when running ossim-cmm or
> ossim-orthoigen.
>
> On Sat, May 12, 2018, 11:28 David Burken, <db...@gm...> wrote:
>
>> Hi Guilaume,
>>
>> I can bring up the file fine. No projection but I can view. No geographic
>> tags in there. So that points to your libtiff maybe?
>>
>> Hope that helps,
>> Dave
>>
>>
>> On Fri, May 11, 2018 at 12:38 PM, Guillaume Drolet <
>> dro...@gm...> wrote:
>>
>>>
>>> Hi,
>>>
>>> I fixed my problem by creating a new version of my input tif file using
>>> gdal_translate.
>>>
>>> I don't know what was causing the error but I suspect OSSIM doesn't like
>>> compressed tiff files: input file had COMPRESSION=YCbCr JPEG prior to
>>> running gdal_translate (no compression in the resulting file).
>>>
>>> Can anyone confirm this?
>>>
>>>
>>>
>>>
>>> *[OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read
>>> Error! t*
>>> From: Guillaume Drolet <droletguillaume@gm...> - 2018-05-10 15:40:48
>>>
>>>  Hello Everyone,
>>>
>>> I few years ago I used ossim-orthoigen for orthorectification of aerial
>>> photographs and I got great help from your list
>>> (http://osgeo-org.1560.x6.nabble.com/Orthorectification-of-
>>> UltraCAM-D-images-td5063637.html)
>>>
>>> I recently re-installed OSSIM version 1.8.19 from Windows binaries I
>>> found
>>> on the web (ossim_x64_1_8_19-20150428-2.exe) to orthorectify new images
>>> from the same camera. My elevation setup is ok and I tested
>>> ossim-orthoigen
>>> on the same image (.TIF) and geometry file I used 4 years ago. But now I
>>> get this error message (excerpt from tracing ossimTiffTileSource):
>>>
>>> ossimSource::print:
>>> theEnableFlag:       1
>>> theInitializedFlag:  0
>>> ossimErrorStatusInterface::print
>>> theErrorStatus:         0
>>> theErrorStatus string:  OSSIM_OK
>>> ossimTiffTileSource::allocateBuffer DEBUG:
>>> buffer_size:  88176
>>> ossimTiffTileSource::allocateBuffer DEBUG:
>>> buffer_size:  262144
>>> ossimTiffTileSource::loadFromRgbaTile Read Error!
>>> Returning error...
>>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
>>> buffer. Return status = false...
>>> ossimTiffTileSource::loadFromRgbaTile Read Error!
>>> Returning error...
>>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
>>> buffer. Return status = false...
>>> ossimTiffTileSource::loadFromRgbaTile Read Error!
>>> Returning error...
>>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
>>> buffer. Return status = false...
>>> ossimTiffTileSource::loadFromRgbaTile Read Error!
>>> Returning error...
>>>
>>> I skimmed the source code but don't have a setup to perform debugging.
>>> Do
>>> you have any ideas what could be the source of this error?
>>>
>>> If this can be helpful, here are the first few lines of the log file
>>> (--ossim-logfile):
>>>
>>> ossimTiffTileSource::open Entered...
>>> File:  JD07137_12007.TIF
>>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
>>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
>>> theMinSampleValue:  -1.#QNAN
>>> theMaxSampleValue:  -1.#QNAN
>>> ossimTiffTileSource::open Debug:image_file:
>>> JD07137_12007.TIF
>>> samples_per_pixel:           3
>>> bits_per_sample:             8
>>> sample_format_unit:          0
>>> min_sample_value:            1
>>> max_sample_value:            255
>>> null_sample_value:           0
>>> theNumberOfDirectories:      7
>>> r0_is_full_res:              1
>>> directory[0]
>>> image width:     11500
>>> image_length:    7500
>>> read method:     READ_RGBA_U8_TILE
>>> planar:          1
>>> photometric:     6
>>> tile_width:      256
>>> tile_length:     256
>>>
>>> directory[1]
>>> image width:     5750
>>> image_length:    3750
>>> read method:     READ_RGBA_U8_TILE
>>> planar:          1
>>> photometric:     6
>>> tile_width:      256
>>> tile_length:     256
>>>
>>> directory[2]
>>> image width:     2875
>>> image_length:    1875
>>> read method:     READ_RGBA_U8_TILE
>>> planar:          1
>>> photometric:     6
>>> tile_width:      256
>>> tile_length:     256
>>>
>>> ...
>>>
>>> ossimSource::print:
>>> theEnableFlag:       1
>>> theInitializedFlag:  0
>>> ossimErrorStatusInterface::print
>>> theErrorStatus:         0
>>> theErrorStatus string:  OSSIM_OK
>>> ossimTiffTileSource::open Entered...
>>> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
>>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
>>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
>>> theMinSampleValue:  -1.#QNAN
>>> theMaxSampleValue:  -1.#QNAN
>>> ossimTiffTileSource::open Debug:image_file:
>>> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
>>> samples_per_pixel:           1
>>> bits_per_sample:             32
>>> sample_format_unit:          3
>>> min_sample_value:            -8.38861e+006
>>> max_sample_value:            8.38861e+006
>>> null_sample_value:           -99999
>>> theNumberOfDirectories:      1
>>> r0_is_full_res:              1
>>> directory[0]
>>> image width:     11022
>>> image_length:    9833
>>> read method:     READ_SCAN_LINE
>>> planar:          1
>>> photometric:     1
>>> rows per strip:  1
>>> tile_width:      256
>>> tile_length:     64
>>>
>>> ossimSource::print:
>>> theEnableFlag:       1
>>> theInitializedFlag:  0
>>> ossimErrorStatusInterface::print
>>> theErrorStatus:         0
>>> theErrorStatus string:  OSSIM_OK
>>> ossimTiffTileSource::open Entered...
>>> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
>>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
>>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
>>> theMinSampleValue:  -1.#QNAN
>>> theMaxSampleValue:  -1.#QNAN
>>> ossimTiffTileSource::open Debug:image_file:
>>> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
>>> samples_per_pixel:           1
>>> bits_per_sample:             32
>>> sample_format_unit:          3
>>> min_sample_value:            -8.38861e+006
>>> max_sample_value:            8.38861e+006
>>> null_sample_value:           -99999
>>> theNumberOfDirectories:      1
>>> r0_is_full_res:              1
>>> directory[0]
>>> image width:     11022
>>> image_length:    9833
>>> read method:     READ_SCAN_LINE
>>> planar:          1
>>> photometric:     1
>>> rows per strip:  1
>>> tile_width:      256
>>> tile_length:     64
>>>
>>> ...
>>>
>>>
>>> I will appreciate any advice on the possible source(s) of this error.
>>>
>>> Best regards,
>>> Guillaume
>>>
>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Check out the vibrant tech community on one of the world's most
>>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>>> _______________________________________________
>>> www.ossim.org
>>> Ossim-developer mailing list
>>> Oss...@li...
>>> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>>>
>>>
>>
>

Re: [OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read Error! t

[Guillaume D. <dro...@gm...>]

Did you try, for example, running ossim-cmm on the image?

ossim-info executes fine but I get the error when running ossim-cmm or
ossim-orthoigen.

On Sat, May 12, 2018, 11:28 David Burken, <db...@gm...> wrote:

> Hi Guilaume,
>
> I can bring up the file fine. No projection but I can view. No geographic
> tags in there. So that points to your libtiff maybe?
>
> Hope that helps,
> Dave
>
>
> On Fri, May 11, 2018 at 12:38 PM, Guillaume Drolet <
> dro...@gm...> wrote:
>
>>
>> Hi,
>>
>> I fixed my problem by creating a new version of my input tif file using
>> gdal_translate.
>>
>> I don't know what was causing the error but I suspect OSSIM doesn't like
>> compressed tiff files: input file had COMPRESSION=YCbCr JPEG prior to
>> running gdal_translate (no compression in the resulting file).
>>
>> Can anyone confirm this?
>>
>>
>>
>>
>> *[OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read
>> Error! t*
>> From: Guillaume Drolet <droletguillaume@gm...> - 2018-05-10 15:40:48
>>
>>  Hello Everyone,
>>
>> I few years ago I used ossim-orthoigen for orthorectification of aerial
>> photographs and I got great help from your list (http://osgeo-org.1560.x6.nabble.com/Orthorectification-of-
>> UltraCAM-D-images-td5063637.html)
>>
>> I recently re-installed OSSIM version 1.8.19 from Windows binaries I found
>> on the web (ossim_x64_1_8_19-20150428-2.exe) to orthorectify new images
>> from the same camera. My elevation setup is ok and I tested ossim-orthoigen
>> on the same image (.TIF) and geometry file I used 4 years ago. But now I
>> get this error message (excerpt from tracing ossimTiffTileSource):
>>
>> ossimSource::print:
>> theEnableFlag:       1
>> theInitializedFlag:  0
>> ossimErrorStatusInterface::print
>> theErrorStatus:         0
>> theErrorStatus string:  OSSIM_OK
>> ossimTiffTileSource::allocateBuffer DEBUG:
>> buffer_size:  88176
>> ossimTiffTileSource::allocateBuffer DEBUG:
>> buffer_size:  262144
>> ossimTiffTileSource::loadFromRgbaTile Read Error!
>> Returning error...
>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
>> buffer. Return status = false...
>> ossimTiffTileSource::loadFromRgbaTile Read Error!
>> Returning error...
>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
>> buffer. Return status = false...
>> ossimTiffTileSource::loadFromRgbaTile Read Error!
>> Returning error...
>> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
>> buffer. Return status = false...
>> ossimTiffTileSource::loadFromRgbaTile Read Error!
>> Returning error...
>>
>> I skimmed the source code but don't have a setup to perform debugging. Do
>> you have any ideas what could be the source of this error?
>>
>> If this can be helpful, here are the first few lines of the log file
>> (--ossim-logfile):
>>
>> ossimTiffTileSource::open Entered...
>> File:  JD07137_12007.TIF
>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
>> theMinSampleValue:  -1.#QNAN
>> theMaxSampleValue:  -1.#QNAN
>> ossimTiffTileSource::open Debug:image_file:
>> JD07137_12007.TIF
>> samples_per_pixel:           3
>> bits_per_sample:             8
>> sample_format_unit:          0
>> min_sample_value:            1
>> max_sample_value:            255
>> null_sample_value:           0
>> theNumberOfDirectories:      7
>> r0_is_full_res:              1
>> directory[0]
>> image width:     11500
>> image_length:    7500
>> read method:     READ_RGBA_U8_TILE
>> planar:          1
>> photometric:     6
>> tile_width:      256
>> tile_length:     256
>>
>> directory[1]
>> image width:     5750
>> image_length:    3750
>> read method:     READ_RGBA_U8_TILE
>> planar:          1
>> photometric:     6
>> tile_width:      256
>> tile_length:     256
>>
>> directory[2]
>> image width:     2875
>> image_length:    1875
>> read method:     READ_RGBA_U8_TILE
>> planar:          1
>> photometric:     6
>> tile_width:      256
>> tile_length:     256
>>
>> ...
>>
>> ossimSource::print:
>> theEnableFlag:       1
>> theInitializedFlag:  0
>> ossimErrorStatusInterface::print
>> theErrorStatus:         0
>> theErrorStatus string:  OSSIM_OK
>> ossimTiffTileSource::open Entered...
>> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
>> theMinSampleValue:  -1.#QNAN
>> theMaxSampleValue:  -1.#QNAN
>> ossimTiffTileSource::open Debug:image_file:
>> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
>> samples_per_pixel:           1
>> bits_per_sample:             32
>> sample_format_unit:          3
>> min_sample_value:            -8.38861e+006
>> max_sample_value:            8.38861e+006
>> null_sample_value:           -99999
>> theNumberOfDirectories:      1
>> r0_is_full_res:              1
>> directory[0]
>> image width:     11022
>> image_length:    9833
>> read method:     READ_SCAN_LINE
>> planar:          1
>> photometric:     1
>> rows per strip:  1
>> tile_width:      256
>> tile_length:     64
>>
>> ossimSource::print:
>> theEnableFlag:       1
>> theInitializedFlag:  0
>> ossimErrorStatusInterface::print
>> theErrorStatus:         0
>> theErrorStatus string:  OSSIM_OK
>> ossimTiffTileSource::open Entered...
>> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
>> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
>> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
>> theMinSampleValue:  -1.#QNAN
>> theMaxSampleValue:  -1.#QNAN
>> ossimTiffTileSource::open Debug:image_file:
>> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
>> samples_per_pixel:           1
>> bits_per_sample:             32
>> sample_format_unit:          3
>> min_sample_value:            -8.38861e+006
>> max_sample_value:            8.38861e+006
>> null_sample_value:           -99999
>> theNumberOfDirectories:      1
>> r0_is_full_res:              1
>> directory[0]
>> image width:     11022
>> image_length:    9833
>> read method:     READ_SCAN_LINE
>> planar:          1
>> photometric:     1
>> rows per strip:  1
>> tile_width:      256
>> tile_length:     64
>>
>> ...
>>
>>
>> I will appreciate any advice on the possible source(s) of this error.
>>
>> Best regards,
>> Guillaume
>>
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>> _______________________________________________
>> www.ossim.org
>> Ossim-developer mailing list
>> Oss...@li...
>> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>>
>>
>

Re: [OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read Error! t

[David B. <db...@gm...>]

Hi Guilaume,

I can bring up the file fine. No projection but I can view. No geographic
tags in there. So that points to your libtiff maybe?

Hope that helps,
Dave


On Fri, May 11, 2018 at 12:38 PM, Guillaume Drolet <
dro...@gm...> wrote:

>
> Hi,
>
> I fixed my problem by creating a new version of my input tif file using
> gdal_translate.
>
> I don't know what was causing the error but I suspect OSSIM doesn't like
> compressed tiff files: input file had COMPRESSION=YCbCr JPEG prior to
> running gdal_translate (no compression in the resulting file).
>
> Can anyone confirm this?
>
>
>
>
> *[OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read
> Error! t*
> From: Guillaume Drolet <droletguillaume@gm...> - 2018-05-10 15:40:48
>
>  Hello Everyone,
>
> I few years ago I used ossim-orthoigen for orthorectification of aerial
> photographs and I got great help from your list (http://osgeo-org.1560.x6.nabble.com/Orthorectification-of-
> UltraCAM-D-images-td5063637.html)
>
> I recently re-installed OSSIM version 1.8.19 from Windows binaries I found
> on the web (ossim_x64_1_8_19-20150428-2.exe) to orthorectify new images
> from the same camera. My elevation setup is ok and I tested ossim-orthoigen
> on the same image (.TIF) and geometry file I used 4 years ago. But now I
> get this error message (excerpt from tracing ossimTiffTileSource):
>
> ossimSource::print:
> theEnableFlag:       1
> theInitializedFlag:  0
> ossimErrorStatusInterface::print
> theErrorStatus:         0
> theErrorStatus string:  OSSIM_OK
> ossimTiffTileSource::allocateBuffer DEBUG:
> buffer_size:  88176
> ossimTiffTileSource::allocateBuffer DEBUG:
> buffer_size:  262144
> ossimTiffTileSource::loadFromRgbaTile Read Error!
> Returning error...
> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> buffer. Return status = false...
> ossimTiffTileSource::loadFromRgbaTile Read Error!
> Returning error...
> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> buffer. Return status = false...
> ossimTiffTileSource::loadFromRgbaTile Read Error!
> Returning error...
> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> buffer. Return status = false...
> ossimTiffTileSource::loadFromRgbaTile Read Error!
> Returning error...
>
> I skimmed the source code but don't have a setup to perform debugging. Do
> you have any ideas what could be the source of this error?
>
> If this can be helpful, here are the first few lines of the log file
> (--ossim-logfile):
>
> ossimTiffTileSource::open Entered...
> File:  JD07137_12007.TIF
> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> theMinSampleValue:  -1.#QNAN
> theMaxSampleValue:  -1.#QNAN
> ossimTiffTileSource::open Debug:image_file:
> JD07137_12007.TIF
> samples_per_pixel:           3
> bits_per_sample:             8
> sample_format_unit:          0
> min_sample_value:            1
> max_sample_value:            255
> null_sample_value:           0
> theNumberOfDirectories:      7
> r0_is_full_res:              1
> directory[0]
> image width:     11500
> image_length:    7500
> read method:     READ_RGBA_U8_TILE
> planar:          1
> photometric:     6
> tile_width:      256
> tile_length:     256
>
> directory[1]
> image width:     5750
> image_length:    3750
> read method:     READ_RGBA_U8_TILE
> planar:          1
> photometric:     6
> tile_width:      256
> tile_length:     256
>
> directory[2]
> image width:     2875
> image_length:    1875
> read method:     READ_RGBA_U8_TILE
> planar:          1
> photometric:     6
> tile_width:      256
> tile_length:     256
>
> ...
>
> ossimSource::print:
> theEnableFlag:       1
> theInitializedFlag:  0
> ossimErrorStatusInterface::print
> theErrorStatus:         0
> theErrorStatus string:  OSSIM_OK
> ossimTiffTileSource::open Entered...
> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> theMinSampleValue:  -1.#QNAN
> theMaxSampleValue:  -1.#QNAN
> ossimTiffTileSource::open Debug:image_file:
> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> samples_per_pixel:           1
> bits_per_sample:             32
> sample_format_unit:          3
> min_sample_value:            -8.38861e+006
> max_sample_value:            8.38861e+006
> null_sample_value:           -99999
> theNumberOfDirectories:      1
> r0_is_full_res:              1
> directory[0]
> image width:     11022
> image_length:    9833
> read method:     READ_SCAN_LINE
> planar:          1
> photometric:     1
> rows per strip:  1
> tile_width:      256
> tile_length:     64
>
> ossimSource::print:
> theEnableFlag:       1
> theInitializedFlag:  0
> ossimErrorStatusInterface::print
> theErrorStatus:         0
> theErrorStatus string:  OSSIM_OK
> ossimTiffTileSource::open Entered...
> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> theMinSampleValue:  -1.#QNAN
> theMaxSampleValue:  -1.#QNAN
> ossimTiffTileSource::open Debug:image_file:
> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> samples_per_pixel:           1
> bits_per_sample:             32
> sample_format_unit:          3
> min_sample_value:            -8.38861e+006
> max_sample_value:            8.38861e+006
> null_sample_value:           -99999
> theNumberOfDirectories:      1
> r0_is_full_res:              1
> directory[0]
> image width:     11022
> image_length:    9833
> read method:     READ_SCAN_LINE
> planar:          1
> photometric:     1
> rows per strip:  1
> tile_width:      256
> tile_length:     64
>
> ...
>
>
> I will appreciate any advice on the possible source(s) of this error.
>
> Best regards,
> Guillaume
>
>
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> Oss...@li...
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>
>

Re: [OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read Error! t

[David B. <db...@gm...>]

Hi, yes it could be that. If you can share a sample we could check it out.
Assuming the underlying libtiff has support for you compression. Take care
Dave


On Fri, May 11, 2018, 9:39 AM Guillaume Drolet <dro...@gm...>
wrote:

>
> Hi,
>
> I fixed my problem by creating a new version of my input tif file using
> gdal_translate.
>
> I don't know what was causing the error but I suspect OSSIM doesn't like
> compressed tiff files: input file had COMPRESSION=YCbCr JPEG prior to
> running gdal_translate (no compression in the resulting file).
>
> Can anyone confirm this?
>
>
>
>
> *[OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read
> Error! t*
> From: Guillaume Drolet <droletguillaume@gm...> - 2018-05-10 15:40:48
>
>  Hello Everyone,
>
> I few years ago I used ossim-orthoigen for orthorectification of aerial
> photographs and I got great help from your list (http://osgeo-org.1560.x6.nabble.com/Orthorectification-of-
> UltraCAM-D-images-td5063637.html)
>
> I recently re-installed OSSIM version 1.8.19 from Windows binaries I found
> on the web (ossim_x64_1_8_19-20150428-2.exe) to orthorectify new images
> from the same camera. My elevation setup is ok and I tested ossim-orthoigen
> on the same image (.TIF) and geometry file I used 4 years ago. But now I
> get this error message (excerpt from tracing ossimTiffTileSource):
>
> ossimSource::print:
> theEnableFlag:       1
> theInitializedFlag:  0
> ossimErrorStatusInterface::print
> theErrorStatus:         0
> theErrorStatus string:  OSSIM_OK
> ossimTiffTileSource::allocateBuffer DEBUG:
> buffer_size:  88176
> ossimTiffTileSource::allocateBuffer DEBUG:
> buffer_size:  262144
> ossimTiffTileSource::loadFromRgbaTile Read Error!
> Returning error...
> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> buffer. Return status = false...
> ossimTiffTileSource::loadFromRgbaTile Read Error!
> Returning error...
> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> buffer. Return status = false...
> ossimTiffTileSource::loadFromRgbaTile Read Error!
> Returning error...
> ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
> buffer. Return status = false...
> ossimTiffTileSource::loadFromRgbaTile Read Error!
> Returning error...
>
> I skimmed the source code but don't have a setup to perform debugging. Do
> you have any ideas what could be the source of this error?
>
> If this can be helpful, here are the first few lines of the log file
> (--ossim-logfile):
>
> ossimTiffTileSource::open Entered...
> File:  JD07137_12007.TIF
> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> theMinSampleValue:  -1.#QNAN
> theMaxSampleValue:  -1.#QNAN
> ossimTiffTileSource::open Debug:image_file:
> JD07137_12007.TIF
> samples_per_pixel:           3
> bits_per_sample:             8
> sample_format_unit:          0
> min_sample_value:            1
> max_sample_value:            255
> null_sample_value:           0
> theNumberOfDirectories:      7
> r0_is_full_res:              1
> directory[0]
> image width:     11500
> image_length:    7500
> read method:     READ_RGBA_U8_TILE
> planar:          1
> photometric:     6
> tile_width:      256
> tile_length:     256
>
> directory[1]
> image width:     5750
> image_length:    3750
> read method:     READ_RGBA_U8_TILE
> planar:          1
> photometric:     6
> tile_width:      256
> tile_length:     256
>
> directory[2]
> image width:     2875
> image_length:    1875
> read method:     READ_RGBA_U8_TILE
> planar:          1
> photometric:     6
> tile_width:      256
> tile_length:     256
>
> ...
>
> ossimSource::print:
> theEnableFlag:       1
> theInitializedFlag:  0
> ossimErrorStatusInterface::print
> theErrorStatus:         0
> theErrorStatus string:  OSSIM_OK
> ossimTiffTileSource::open Entered...
> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> theMinSampleValue:  -1.#QNAN
> theMaxSampleValue:  -1.#QNAN
> ossimTiffTileSource::open Debug:image_file:
> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> samples_per_pixel:           1
> bits_per_sample:             32
> sample_format_unit:          3
> min_sample_value:            -8.38861e+006
> max_sample_value:            8.38861e+006
> null_sample_value:           -99999
> theNumberOfDirectories:      1
> r0_is_full_res:              1
> directory[0]
> image width:     11022
> image_length:    9833
> read method:     READ_SCAN_LINE
> planar:          1
> photometric:     1
> rows per strip:  1
> tile_width:      256
> tile_length:     64
>
> ossimSource::print:
> theEnableFlag:       1
> theInitializedFlag:  0
> ossimErrorStatusInterface::print
> theErrorStatus:         0
> theErrorStatus string:  OSSIM_OK
> ossimTiffTileSource::open Entered...
> File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
> sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
> theMinSampleValue:  -1.#QNAN
> theMaxSampleValue:  -1.#QNAN
> ossimTiffTileSource::open Debug:image_file:
> E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
> samples_per_pixel:           1
> bits_per_sample:             32
> sample_format_unit:          3
> min_sample_value:            -8.38861e+006
> max_sample_value:            8.38861e+006
> null_sample_value:           -99999
> theNumberOfDirectories:      1
> r0_is_full_res:              1
> directory[0]
> image width:     11022
> image_length:    9833
> read method:     READ_SCAN_LINE
> planar:          1
> photometric:     1
> rows per strip:  1
> tile_width:      256
> tile_length:     64
>
> ...
>
>
> I will appreciate any advice on the possible source(s) of this error.
>
> Best regards,
> Guillaume
>
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> Oss...@li...
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>

Re: [OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read Error! t

[Guillaume D. <dro...@gm...>]

Hi,

I fixed my problem by creating a new version of my input tif file using
gdal_translate.

I don't know what was causing the error but I suspect OSSIM doesn't like
compressed tiff files: input file had COMPRESSION=YCbCr JPEG prior to
running gdal_translate (no compression in the resulting file).

Can anyone confirm this?




*[OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read Error!
t*
From: Guillaume Drolet <droletguillaume@gm...> - 2018-05-10 15:40:48

 Hello Everyone,

I few years ago I used ossim-orthoigen for orthorectification of aerial
photographs and I got great help from your list
(http://osgeo-org.1560.x6.nabble.com/Orthorectification-of-
UltraCAM-D-images-td5063637.html)

I recently re-installed OSSIM version 1.8.19 from Windows binaries I found
on the web (ossim_x64_1_8_19-20150428-2.exe) to orthorectify new images
from the same camera. My elevation setup is ok and I tested ossim-orthoigen
on the same image (.TIF) and geometry file I used 4 years ago. But now I
get this error message (excerpt from tracing ossimTiffTileSource):

ossimSource::print:
theEnableFlag:       1
theInitializedFlag:  0
ossimErrorStatusInterface::print
theErrorStatus:         0
theErrorStatus string:  OSSIM_OK
ossimTiffTileSource::allocateBuffer DEBUG:
buffer_size:  88176
ossimTiffTileSource::allocateBuffer DEBUG:
buffer_size:  262144
ossimTiffTileSource::loadFromRgbaTile Read Error!
Returning error...
ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
buffer. Return status = false...
ossimTiffTileSource::loadFromRgbaTile Read Error!
Returning error...
ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
buffer. Return status = false...
ossimTiffTileSource::loadFromRgbaTile Read Error!
Returning error...
ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
buffer. Return status = false...
ossimTiffTileSource::loadFromRgbaTile Read Error!
Returning error...

I skimmed the source code but don't have a setup to perform debugging. Do
you have any ideas what could be the source of this error?

If this can be helpful, here are the first few lines of the log file
(--ossim-logfile):

ossimTiffTileSource::open Entered...
File:  JD07137_12007.TIF
ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
theMinSampleValue:  -1.#QNAN
theMaxSampleValue:  -1.#QNAN
ossimTiffTileSource::open Debug:image_file:
JD07137_12007.TIF
samples_per_pixel:           3
bits_per_sample:             8
sample_format_unit:          0
min_sample_value:            1
max_sample_value:            255
null_sample_value:           0
theNumberOfDirectories:      7
r0_is_full_res:              1
directory[0]
image width:     11500
image_length:    7500
read method:     READ_RGBA_U8_TILE
planar:          1
photometric:     6
tile_width:      256
tile_length:     256

directory[1]
image width:     5750
image_length:    3750
read method:     READ_RGBA_U8_TILE
planar:          1
photometric:     6
tile_width:      256
tile_length:     256

directory[2]
image width:     2875
image_length:    1875
read method:     READ_RGBA_U8_TILE
planar:          1
photometric:     6
tile_width:      256
tile_length:     256

...

ossimSource::print:
theEnableFlag:       1
theInitializedFlag:  0
ossimErrorStatusInterface::print
theErrorStatus:         0
theErrorStatus string:  OSSIM_OK
ossimTiffTileSource::open Entered...
File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
theMinSampleValue:  -1.#QNAN
theMaxSampleValue:  -1.#QNAN
ossimTiffTileSource::open Debug:image_file:
E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
samples_per_pixel:           1
bits_per_sample:             32
sample_format_unit:          3
min_sample_value:            -8.38861e+006
max_sample_value:            8.38861e+006
null_sample_value:           -99999
theNumberOfDirectories:      1
r0_is_full_res:              1
directory[0]
image width:     11022
image_length:    9833
read method:     READ_SCAN_LINE
planar:          1
photometric:     1
rows per strip:  1
tile_width:      256
tile_length:     64

ossimSource::print:
theEnableFlag:       1
theInitializedFlag:  0
ossimErrorStatusInterface::print
theErrorStatus:         0
theErrorStatus string:  OSSIM_OK
ossimTiffTileSource::open Entered...
File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
theMinSampleValue:  -1.#QNAN
theMaxSampleValue:  -1.#QNAN
ossimTiffTileSource::open Debug:image_file:
E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
samples_per_pixel:           1
bits_per_sample:             32
sample_format_unit:          3
min_sample_value:            -8.38861e+006
max_sample_value:            8.38861e+006
null_sample_value:           -99999
theNumberOfDirectories:      1
r0_is_full_res:              1
directory[0]
image width:     11022
image_length:    9833
read method:     READ_SCAN_LINE
planar:          1
photometric:     1
rows per strip:  1
tile_width:      256
tile_length:     64

...


I will appreciate any advice on the possible source(s) of this error.

Best regards,
Guillaume

[OSSIM] orthoigen error: ossimTiffTileSource::loadFromRgbaTile Read Error! t

[Guillaume D. <dro...@gm...>]

 Hello Everyone,

I few years ago I used ossim-orthoigen for orthorectification of aerial
photographs and I got great help from your list (
http://osgeo-org.1560.x6.nabble.com/Orthorectification-of-
UltraCAM-D-images-td5063637.html)

I recently re-installed OSSIM version 1.8.19 from Windows binaries I found
on the web (ossim_x64_1_8_19-20150428-2.exe) to orthorectify new images
from the same camera. My elevation setup is ok and I tested ossim-orthoigen
on the same image (.TIF) and geometry file I used 4 years ago. But now I
get this error message (excerpt from tracing ossimTiffTileSource):

ossimSource::print:
theEnableFlag:       1
theInitializedFlag:  0
ossimErrorStatusInterface::print
theErrorStatus:         0
theErrorStatus string:  OSSIM_OK
ossimTiffTileSource::allocateBuffer DEBUG:
buffer_size:  88176
ossimTiffTileSource::allocateBuffer DEBUG:
buffer_size:  262144
ossimTiffTileSource::loadFromRgbaTile Read Error!
Returning error...
ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
buffer. Return status = false...
ossimTiffTileSource::loadFromRgbaTile Read Error!
Returning error...
ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
buffer. Return status = false...
ossimTiffTileSource::loadFromRgbaTile Read Error!
Returning error...
ossimTiffTileSource::getTile(ossimImageData*, resLevel) Error filling
buffer. Return status = false...
ossimTiffTileSource::loadFromRgbaTile Read Error!
Returning error...

I skimmed the source code but don't have a setup to perform debugging. Do
you have any ideas what could be the source of this error?

If this can be helpful, here are the first few lines of the log file
(--ossim-logfile):

ossimTiffTileSource::open Entered...
File:  JD07137_12007.TIF
ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
theMinSampleValue:  -1.#QNAN
theMaxSampleValue:  -1.#QNAN
ossimTiffTileSource::open Debug:image_file:
JD07137_12007.TIF
samples_per_pixel:           3
bits_per_sample:             8
sample_format_unit:          0
min_sample_value:            1
max_sample_value:            255
null_sample_value:           0
theNumberOfDirectories:      7
r0_is_full_res:              1
directory[0]
image width:     11500
image_length:    7500
read method:     READ_RGBA_U8_TILE
planar:          1
photometric:     6
tile_width:      256
tile_length:     256

directory[1]
image width:     5750
image_length:    3750
read method:     READ_RGBA_U8_TILE
planar:          1
photometric:     6
tile_width:      256
tile_length:     256

directory[2]
image width:     2875
image_length:    1875
read method:     READ_RGBA_U8_TILE
planar:          1
photometric:     6
tile_width:      256
tile_length:     256

...

ossimSource::print:
theEnableFlag:       1
theInitializedFlag:  0
ossimErrorStatusInterface::print
theErrorStatus:         0
theErrorStatus string:  OSSIM_OK
ossimTiffTileSource::open Entered...
File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
theMinSampleValue:  -1.#QNAN
theMaxSampleValue:  -1.#QNAN
ossimTiffTileSource::open Debug:image_file:
E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
samples_per_pixel:           1
bits_per_sample:             32
sample_format_unit:          3
min_sample_value:            -8.38861e+006
max_sample_value:            8.38861e+006
null_sample_value:           -99999
theNumberOfDirectories:      1
r0_is_full_res:              1
directory[0]
image width:     11022
image_length:    9833
read method:     READ_SCAN_LINE
planar:          1
photometric:     1
rows per strip:  1
tile_width:      256
tile_length:     64

ossimSource::print:
theEnableFlag:       1
theInitializedFlag:  0
ossimErrorStatusInterface::print
theErrorStatus:         0
theErrorStatus string:  OSSIM_OK
ossimTiffTileSource::open Entered...
File:  E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
ossimTiffTileSource::open ("D:\dev\vs11_x64\ossim\src\os
sim\imaging\ossimTiffTileSource.cpp", line 547) DEBUG: DEBUG:
theMinSampleValue:  -1.#QNAN
theMaxSampleValue:  -1.#QNAN
ossimTiffTileSource::open Debug:image_file:
E:\Data\Spatial\OSSIM\elevation\srtm_v3_1arc\0000_0000.tif
samples_per_pixel:           1
bits_per_sample:             32
sample_format_unit:          3
min_sample_value:            -8.38861e+006
max_sample_value:            8.38861e+006
null_sample_value:           -99999
theNumberOfDirectories:      1
r0_is_full_res:              1
directory[0]
image width:     11022
image_length:    9833
read method:     READ_SCAN_LINE
planar:          1
photometric:     1
rows per strip:  1
tile_width:      256
tile_length:     64

...


I will appreciate any advice on the possible source(s) of this error.

Best regards,
Guillaume

Re: [OSSIM] Implementing orthoigen for all image projection

[Oscar K. <osc...@ya...>]

 The input projection is created automatically when you load your image file, assuming you image file is of a common, known format. This is complicated code that makes use of the ossimProjectionFactoryRegistry to instantiate the proper projection or sensor model given the image metadata. I cannot explain all the details. I'm not even sure what you are trying to do. You'll have to look at that code and figure it out from there. I also sent you the link to the sensor model document.
Are you referring perhaps to the output projection specification? That is, the projection you are orthorectifying to. That is an option on ossim-orthoigen. Run the command with --help to get the full usage.

    On Monday, March 19, 2018, 12:49:31 AM EDT, vedantkolarkar <ved...@gm...> wrote:  
 
 Orthoigen seems to pick up RPB file only for quickbirdrpc type projection.
What changes do i have to make in orthoigen related programs, so that it
will pick up all general image projection type? Is their a specific tag that
I have to change so that it will work for all image projection types??
Conversion from RPC to RPB can now be done easily as per your
guidelines.......just want to implement orthoigen for any image projection.



--
Sent from: http://osgeo-org.1560.x6.nabble.com/Ossim-developer-f3764496.html

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
www.ossim.org
Ossim-developer mailing list
Oss...@li...
https://lists.sourceforge.net/lists/listinfo/ossim-developer
  

[OSSIM] Implementing orthoigen for all image projection

[vedantkolarkar <ved...@gm...>]

Orthoigen seems to pick up RPB file only for quickbirdrpc type projection.
What changes do i have to make in orthoigen related programs, so that it
will pick up all general image projection type? Is their a specific tag that
I have to change so that it will work for all image projection types??
Conversion from RPC to RPB can now be done easily as per your
guidelines.......just want to implement orthoigen for any image projection.



--
Sent from: http://osgeo-org.1560.x6.nabble.com/Ossim-developer-f3764496.html

Re: [OSSIM] How can i make use of sensor model for geometric correction???

[Oscar K. <osc...@ra...>]

When you implement the sensor model, including the automatic detection of
the metadata through the sensor model factory system, orthoigen will
automatically include your sensor model when you load the image. Just make
sure your model factory is registered with the
ossimProjectionFactoryRegistry, I assume the image is in a supported format
(TIFF?)

When orthoigen loads an image, part of the open is to query the
ossimProjectionFactoryRegistry for a projection (specifically a sensor
model in your case). If your model can successfully instantiate itself from
the available metadata on disk, the projection registry will return that
instance to orthoigen. You just need to make sure that there isn't another
model (RPC for example) that will find its own metadata in the TIFF or
separate file (e.g., a .RPB file), because it would be returned before your
model. You can add your model to the front of the registry to make sure
your factory has the first opportunity to create your model.

Look at the existing code and follow the execution using your IDE. You can
run "ossim-info -p <image-filename>". This will exercise the factory code.

Oscar



<http://www.radiantsolutions.com/>

Oscar L Kramer
Senior Software Engineer
+1.863.703.4535 office <//+1.863.703.4535>
osc...@ra...

On Thu, Mar 1, 2018 at 1:46 AM, Vedant Kolarkar <ved...@gm...>
wrote:

> Sir,
>      I want to implement a new model for CARTOSAT sensor. Previously I
> used ossim's quickbirdrpc model for orthorectification using orthoigen. Now
> i want to do the same for cartosat images. I'll go through the link you
> shared and try to create a new class. After creating a new model, is it
> possibe to interface orthoigen with it??  Thank you for your help.
>
> On Wed, Feb 28, 2018 at 9:24 PM, Oscar Kramer <oscar.kramer@
> radiantsolutions.com> wrote:
>
>> Sorry for all the repeat emails. I'm having trouble with the list server.
>>
>>
>> <http://www.radiantsolutions.com/>
>>
>> Oscar L Kramer
>> Senior Software Engineer
>> +1.863.703.4535 office <//+1.863.703.4535>
>> osc...@ra...
>>
>> On Wed, Feb 28, 2018 at 12:30 AM, vedantkolarkar <
>> ved...@gm...> wrote:
>>
>>> Is it possible to use ossim-orthoigen in sensor modelling, if not can i
>>> make
>>> a new sensor model and incorporate orthoigen in it??
>>>
>>>
>>>
>>> --
>>> Sent from: http://osgeo-org.1560.x6.nabble.com/Ossim-developer-f3764496
>>> .html
>>>
>>> ------------------------------------------------------------
>>> ------------------
>>> Check out the vibrant tech community on one of the world's most
>>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>>> _______________________________________________
>>> www.ossim.org
>>> Ossim-developer mailing list
>>> Oss...@li...
>>> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>>>
>>>
>>
>>
>>
>> This electronic communication and any attachments may contain
>> confidential and proprietary information of DigitalGlobe, Inc. If you are
>> not the intended recipient, or an agent or employee responsible for
>> delivering this communication to the intended recipient, or if you have
>> received this communication in error, please do not print, copy,
>> retransmit, disseminate or otherwise use the information. Please indicate
>> to the sender that you have received this communication in error, and
>> delete the copy you received.
>>
>> DigitalGlobe reserves the right to monitor any electronic communication
>> sent or received by its employees, agents or representatives.
>>
>>
>

-- 


This electronic communication and any attachments may contain confidential 
and proprietary information of DigitalGlobe, Inc. If you are not the 
intended recipient, or an agent or employee responsible for delivering this 
communication to the intended recipient, or if you have received this 
communication in error, please do not print, copy, retransmit, disseminate 
or otherwise use the information. Please indicate to the sender that you 
have received this communication in error, and delete the copy you received.

DigitalGlobe reserves the right to monitor any electronic communication 
sent or received by its employees, agents or representatives.

Re: [OSSIM] How can i make use of sensor model for geometric correction???

[Oscar K. <osc...@ra...>]

V,

A sensor model is used *in* orthoigen. Orthoigen is an application. A
sensor model is a piece of code (a class) that is referenced by orthoigen
to perform image-to-ground projection and ground-to-image transformation.

Are you working with imagery for which OSSIM already has a sensor model
defined? Or do you need to implement a new model of an unsupported sensor?
If the latter, you will need to create a class derived from
ossimSensorModel. Use an existing model as a guide, for example
ossimFcsiModel.

You can read about sensor modeling in OSSIM here
<http://download.osgeo.org/ossim/docs/pdfs/SensorModeling.pdf>.


Oscar

-- 


This electronic communication and any attachments may contain confidential 
and proprietary information of DigitalGlobe, Inc. If you are not the 
intended recipient, or an agent or employee responsible for delivering this 
communication to the intended recipient, or if you have received this 
communication in error, please do not print, copy, retransmit, disseminate 
or otherwise use the information. Please indicate to the sender that you 
have received this communication in error, and delete the copy you received.

DigitalGlobe reserves the right to monitor any electronic communication 
sent or received by its employees, agents or representatives.

Re: [OSSIM] How can i make use of sensor model for geometric correction???

[Andrew D. J. <ajo...@ke...>]

Custom sensor models can be incorporated as a plugin if for some reason you
don't want to/can't add it into the trunk.
You pretty much need to setup a plugin init to register your shared lib., a
class factory inheriting ossimProjectionFactoryBase, and then build your own
class off of ossimSensorModel.

Take a look at the csm plugin:
https://github.com/ossimlabs/ossim-csm-plugin

-Andy

-----Original Message-----
From: vedantkolarkar [mailto:ved...@gm...] 
Sent: Wednesday, February 28, 2018 12:30 AM
To: oss...@li...
Subject: [OSSIM] How can i make use of sensor model for geometric
correction???

Is it possible to use ossim-orthoigen in sensor modelling, if not can i make
a new sensor model and incorporate orthoigen in it?? 



--
Sent from:
https://urldefense.proofpoint.com/v2/url?u=http-3A__osgeo-2Dorg.1560.x6.nabb
le.com_Ossim-2Ddeveloper-2Df3764496.html&d=DwICAg&c=31nHN1tvZeuWBT6LwDN4Ngk1
qezfsYHyolgGeY2ZhlU&r=7T-tcCZEv1UtWANbvWK5_xMS1HF3pvNdFIgqlwqJo6o&m=gRPXWQDR
KQ6UECk4i2WoibHDGaj50lMVcKS2C1wXVeA&s=LQwmt_QHVXDBWnDFy5BgrnHLSHnRzYnUG8bK7j
diqjs&e=

----------------------------------------------------------------------------
--
Check out the vibrant tech community on one of the world's most engaging
tech sites, Slashdot.org!
https://urldefense.proofpoint.com/v2/url?u=http-3A__sdm.link_slashdot&d=DwIC
Ag&c=31nHN1tvZeuWBT6LwDN4Ngk1qezfsYHyolgGeY2ZhlU&r=7T-tcCZEv1UtWANbvWK5_xMS1
HF3pvNdFIgqlwqJo6o&m=gRPXWQDRKQ6UECk4i2WoibHDGaj50lMVcKS2C1wXVeA&s=z_ie_-_-K
R2zv9beo9qIfJV2qqRJ7iBOJRDDQ7KKAdQ&e=
_______________________________________________
www.ossim.org
Ossim-developer mailing list
Oss...@li...
https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.sourceforge.net_l
ists_listinfo_ossim-2Ddeveloper&d=DwICAg&c=31nHN1tvZeuWBT6LwDN4Ngk1qezfsYHyo
lgGeY2ZhlU&r=7T-tcCZEv1UtWANbvWK5_xMS1HF3pvNdFIgqlwqJo6o&m=gRPXWQDRKQ6UECk4i
2WoibHDGaj50lMVcKS2C1wXVeA&s=bzrbI3H1sD01dSw9KF7j-ow-Q1m7Ruu7X0JWLfYOHts&e=

[OSSIM] How can i make use of sensor model for geometric correction???

[vedantkolarkar <ved...@gm...>]

Is it possible to use ossim-orthoigen in sensor modelling, if not can i make
a new sensor model and incorporate orthoigen in it?? 



--
Sent from: http://osgeo-org.1560.x6.nabble.com/Ossim-developer-f3764496.html

Re: [OSSIM] What's up with "degug" ?

[Garrett P. <gar...@gm...>]

Hello Eric:

Fixed in a branch I am working


Take care

Garrett Potts	
> On Feb 10, 2018, at 7:47 AM, Garrett Potts <gar...@gm...> wrote:
> 
> Hello Eric:
> 
> Good catch….Probably my fault everytime I hit git commit it bring s the editor up for the commit message and then I save and I start typing git push and the editor doesn’t close fast enough and so the straggling “g” gets inserted or something into the code.  Aggrevating.
> 
> Will do a search and find all the files and fix the g  to a b.   
> 
> Thank you
> 
> 
> Garrett
> 
>> On Feb 9, 2018, at 6:24 PM, Eric Hirschorn <ehi...@ke... <mailto:ehi...@ke...>> wrote:
>> 
>> Hi ossim developers,
>> 
>> I noticed there are some source files that use the term "degug" where "debug" should appear.  Here's an example:
>> 
>> static ossimTrace traceDebug("ossimJpegTileSource:degug");
>> 
>> Even in the GIT version of ossim core there are 7 source files like this one.  Outside of core, there are many (dozens) more.
>> 
>> Just want to make sure what the intentions of the developers are: are they typo bugs or intentional disables of the debugging mechanism?
>> 
>> Regards,
>> Eric Hirschorn
>> 
>> Software Design Engineer
>> KeyW Corporation | Remote Sensing Division
>> 250 Clark Street | North Andover, MA 01845
>> Direct: 978 396 0399
>> Skype:  eric.hirschorn
>> Email:  ehi...@ke... <mailto:ehi...@ke...>
>> 
>> ------------------------------------------------------------------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, Slashdot.org <http://slashdot.org/>! http://sdm.link/slashdot_______________________________________________ <http://sdm.link/slashdot_______________________________________________>
>> www.ossim.org <http://www.ossim.org/>
>> Ossim-developer mailing list
>> Oss...@li... <mailto:Oss...@li...>
>> https://lists.sourceforge.net/lists/listinfo/ossim-developer <https://lists.sourceforge.net/lists/listinfo/ossim-developer>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot_______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> Oss...@li...
> https://lists.sourceforge.net/lists/listinfo/ossim-developer

Re: [OSSIM] What's up with "degug" ?

[Garrett P. <gar...@gm...>]

Hello Eric:

Good catch….Probably my fault everytime I hit git commit it bring s the editor up for the commit message and then I save and I start typing git push and the editor doesn’t close fast enough and so the straggling “g” gets inserted or something into the code.  Aggrevating.

Will do a search and find all the files and fix the g  to a b.   

Thank you


Garrett

> On Feb 9, 2018, at 6:24 PM, Eric Hirschorn <ehi...@ke...> wrote:
> 
> Hi ossim developers,
> 
> I noticed there are some source files that use the term "degug" where "debug" should appear.  Here's an example:
> 
> static ossimTrace traceDebug("ossimJpegTileSource:degug");
> 
> Even in the GIT version of ossim core there are 7 source files like this one.  Outside of core, there are many (dozens) more.
> 
> Just want to make sure what the intentions of the developers are: are they typo bugs or intentional disables of the debugging mechanism?
> 
> Regards,
> Eric Hirschorn
> 
> Software Design Engineer
> KeyW Corporation | Remote Sensing Division
> 250 Clark Street | North Andover, MA 01845
> Direct: 978 396 0399
> Skype:  eric.hirschorn
> Email:  ehi...@ke... <mailto:ehi...@ke...>
> 
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org <http://slashdot.org/>! http://sdm.link/slashdot_______________________________________________ <http://sdm.link/slashdot_______________________________________________>
> www.ossim.org <http://www.ossim.org/>
> Ossim-developer mailing list
> Oss...@li... <mailto:Oss...@li...>
> https://lists.sourceforge.net/lists/listinfo/ossim-developer <https://lists.sourceforge.net/lists/listinfo/ossim-developer>

[OSSIM] What's up with "degug" ?

[Eric H. <ehi...@ke...>]

Hi ossim developers,

I noticed there are some source files that use the term "degug" where "debug" should appear.  Here's an example:

static ossimTrace traceDebug("ossimJpegTileSource:degug");

Even in the GIT version of ossim core there are 7 source files like this one.  Outside of core, there are many (dozens) more.

Just want to make sure what the intentions of the developers are: are they typo bugs or intentional disables of the debugging mechanism?

Regards,
Eric Hirschorn

Software Design Engineer
KeyW Corporation | Remote Sensing Division
250 Clark Street | North Andover, MA 01845
Direct: 978 396 0399
Skype:  eric.hirschorn
Email:  ehi...@ke...

Re: [OSSIM] RPC text file as an input to ossim-orthoigen with refrence image

[Oscar K. <osc...@di...>]

Sorry forgot to address the second question. How do you envision using
tiepoints for orthorectification? I presume you want to stitch multiple
images together. This involves somehow adjusting the input projections (the
RPCs?) so that the images line up -- typically via a least-squares
triangulation -- before inputting the adjusted projections into orthoigen.
Orthoigen itself wouldn't know what to do with tiepoints. The triangulation
capability may exist in the ossim registration plugin, but I am not
familiar with that code unfortunately.


<http://www.radiantsolutions.com/>

Oscar Kramer
Senior Software Engineer
+1.305.450.2775 office <//+1.305.450.2775>
osc...@di...



On Wed, Jan 10, 2018 at 1:34 AM, vedantkolarkar <ved...@gm...>
wrote:

> Is it possible to give rpc text file as an input to ossim-orthoigen with
> reference image for better rectification?? I tried giving text file as
> input
> but it skipped .txt file. Another question- I have difftiepts.txt which i
> got from ossim-correl....is it possible to give this as an input to
> orthoigen??
>
>
>
> --
> Sent from: http://osgeo-org.1560.x6.nabble.com/Ossim-developer-
> f3764496.html
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> Oss...@li...
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>
>
> This electronic communication and any attachments may contain confidential
> and proprietary information of DigitalGlobe, Inc. If you are not the
> intended recipient, or an agent or employee responsible for delivering this
> communication to the intended recipient, or if you have received this
> communication in error, please do not print, copy, retransmit, disseminate
> or otherwise use the information. Please indicate to the sender that you
> have received this communication in error, and delete the copy you received.
>
> DigitalGlobe reserves the right to monitor any electronic communication
> sent or received by its employees, agents or representatives.
>

-- 


This electronic communication and any attachments may contain confidential 
and proprietary information of DigitalGlobe, Inc. If you are not the 
intended recipient, or an agent or employee responsible for delivering this 
communication to the intended recipient, or if you have received this 
communication in error, please do not print, copy, retransmit, disseminate 
or otherwise use the information. Please indicate to the sender that you 
have received this communication in error, and delete the copy you received.

DigitalGlobe reserves the right to monitor any electronic communication 
sent or received by its employees, agents or representatives.

Re: [OSSIM] RPC text file as an input to ossim-orthoigen with refrence image

[Oscar K. <osc...@di...>]

OSSIM can read RPCs as the input geometry but it needs to be in a specific
format. Are you a coder or just using the applications?

Briefly, the RPCs need to be either embedded in a TIFF (or NITF) tag or as
a separate *.RPB file often found with DigitalGlobe imagery. If you can
convert your text file to the format in *.RPB, OSSIM should pick it up and
set up the RPC as the image projection.

I'm attaching a sample RPB file for you to use as a template. There's a lot
of info online regarding the coefficients, offsets, and scaling parameters
(though it sounds like you already have the RPC definition). Just name the
RPB file the same as the image file but with ".RPB" as the extension, and
OSSIM will load it as an ossimQuickbirdRpcModel. Even though it's named
"quickbird", it will serve for any RPC (in "B" format). If your RPC is type
"A" (the older form), name the file with extension ".RPA".

Good luck.

Oscar


<http://www.radiantsolutions.com/>

Oscar Kramer
Senior Software Engineer
+1.305.450.2775 office <//+1.305.450.2775>
osc...@di...



On Wed, Jan 10, 2018 at 1:34 AM, vedantkolarkar <ved...@gm...>
wrote:

> Is it possible to give rpc text file as an input to ossim-orthoigen with
> reference image for better rectification?? I tried giving text file as
> input
> but it skipped .txt file. Another question- I have difftiepts.txt which i
> got from ossim-correl....is it possible to give this as an input to
> orthoigen??
>
>
>
> --
> Sent from: http://osgeo-org.1560.x6.nabble.com/Ossim-developer-
> f3764496.html
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> Oss...@li...
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>
>
> This electronic communication and any attachments may contain confidential
> and proprietary information of DigitalGlobe, Inc. If you are not the
> intended recipient, or an agent or employee responsible for delivering this
> communication to the intended recipient, or if you have received this
> communication in error, please do not print, copy, retransmit, disseminate
> or otherwise use the information. Please indicate to the sender that you
> have received this communication in error, and delete the copy you received.
>
> DigitalGlobe reserves the right to monitor any electronic communication
> sent or received by its employees, agents or representatives.
>

-- 


This electronic communication and any attachments may contain confidential 
and proprietary information of DigitalGlobe, Inc. If you are not the 
intended recipient, or an agent or employee responsible for delivering this 
communication to the intended recipient, or if you have received this 
communication in error, please do not print, copy, retransmit, disseminate 
or otherwise use the information. Please indicate to the sender that you 
have received this communication in error, and delete the copy you received.

DigitalGlobe reserves the right to monitor any electronic communication 
sent or received by its employees, agents or representatives.

[OSSIM] RPC text file as an input to ossim-orthoigen with refrence image

[vedantkolarkar <ved...@gm...>]

Is it possible to give rpc text file as an input to ossim-orthoigen with
reference image for better rectification?? I tried giving text file as input
but it skipped .txt file. Another question- I have difftiepts.txt which i
got from ossim-correl....is it possible to give this as an input to
orthoigen??



--
Sent from: http://osgeo-org.1560.x6.nabble.com/Ossim-developer-f3764496.html