You can subscribe to this list here.
| 2002 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(5) |
Dec
(15) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2003 |
Jan
(1) |
Feb
(11) |
Mar
(18) |
Apr
(1) |
May
(10) |
Jun
(5) |
Jul
(7) |
Aug
(5) |
Sep
(12) |
Oct
(2) |
Nov
|
Dec
(13) |
| 2004 |
Jan
(13) |
Feb
(46) |
Mar
(20) |
Apr
(10) |
May
(13) |
Jun
(8) |
Jul
(5) |
Aug
(1) |
Sep
|
Oct
|
Nov
|
Dec
|
| 2005 |
Jan
(1) |
Feb
|
Mar
|
Apr
|
May
|
Jun
(1) |
Jul
(47) |
Aug
(14) |
Sep
(7) |
Oct
(1) |
Nov
|
Dec
|
| 2006 |
Jan
(11) |
Feb
(17) |
Mar
(5) |
Apr
(9) |
May
(1) |
Jun
|
Jul
(5) |
Aug
(4) |
Sep
(1) |
Oct
|
Nov
(1) |
Dec
(1) |
| 2007 |
Jan
(5) |
Feb
(2) |
Mar
|
Apr
(1) |
May
|
Jun
(2) |
Jul
(18) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2009 |
Jan
|
Feb
(3) |
Mar
(8) |
Apr
(5) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
| 2010 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
(2) |
Oct
(6) |
Nov
|
Dec
|
| 2011 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
(1) |
Sep
(2) |
Oct
(4) |
Nov
|
Dec
|
| 2012 |
Jan
|
Feb
(2) |
Mar
(4) |
Apr
|
May
|
Jun
(4) |
Jul
(1) |
Aug
(2) |
Sep
|
Oct
(1) |
Nov
(3) |
Dec
|
| 2013 |
Jan
|
Feb
|
Mar
(1) |
Apr
(2) |
May
|
Jun
|
Jul
|
Aug
|
Sep
(3) |
Oct
|
Nov
|
Dec
|
| 2016 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
|
|
From: DRUEL J. <joc...@li...> - 2004-02-06 23:09:45
|
Hi, Here's the final piece to my contributions: a help page for interferometry_gui. I tried to show how to use it with the example pictures that are found in the images/photonics directory. I think, for me, it's all for SIP-0.3.0 Of course, I've still got ideas, but they'll be for the next development serie. Best regards Jocelyn -- Powered by Linux |
|
From: DRUEL J. <joc...@li...> - 2004-02-06 23:05:00
|
Hi, I continue my investigations with the new release: 2 corrections for my part: * unwrapp did not work because a threshold value was still at 127 instead of 0.5: it's corrected. * I tried to test extensively the interferometry_gui.dem: I corrected some things (see the changelog at the beginning of the function). It should work fine. I talk about a help page for interferometry_gui.dem: just give me a bit more time to allow me to write it. I join the 2 files. Best regards Jocelyn -- Powered by Linux |
|
From: DRUEL J. <joc...@li...> - 2004-02-06 19:27:01
|
Hi, As I test SIP-0.3, I try to correct what goes wrong with me, and improve (a little) some things: * imconv.sci: just added the correction syntax in the error message and added a warning that it does not work with colored pictures. * mkfilter.sci: I just added an error message if the number of arguments is incorrect. I changed (a very little) the way to calculate the low-pass filters: with mask=mask/sum(mask): it's a bit more generic. * unwrapl: it was working fine (thanks Ricardo). I added just an indication that the unwrapped phase is not in 0-1 range. I adapted the example of the man page. * improfile: I added a lum_max (optional) parameter: the default is 255 (8bit image). It's only used to display profiles on the screen (because values in the profile are in 0-1 range). I prefer this solution to the one that consists of multiplying the profile by the (horizontal or vertical) size of the picture: it allows to make direct measurement on the screen. This multiplication coefficient is customizable by the user. The man page was rewritten to explain this. * each time, I tried to write an error function that also indicates what the user should have done. * That's all for now (but I'd like to perform some other tests if I have some time: with the interf_demo mainly). I believe SIP-0.3.0 will be a great release. Best regards Jocelyn -- Powered by Linux |
|
From: DRUEL J. <joc...@li...> - 2004-02-06 16:04:59
|
Hi, What a pleasure to have binaries availables: no libraries to install, no upgrade of gcc,... real comfort ! Thanks Ricardo. So the binaries seem to work well for me, on my Scilab-2.7 But, as I tried the function improfile, I discovered something wrong: a1=imread(SIPDIR+'images/ararauna.jpg'); max(a1) gives 1: allright, that's the new way SIP behaves. a2=imread(SIPDIR+'images/onca.gif'); max(a2) gives 256 = problem but a3=gray_imread(SIPDIR+'images/onca.gif'); max(a3) is <1: fine There seems to be only a problem with imread reading indexed images. Something else: I did not manage to compile this rc2 from sources: I join the file: my config=scilab-2.6, ImageMagick-5.5.7, gcc2.96 gcc-3.0 gives the same result. -- Powered by Linux |
|
From: Ricardo F. <rf...@if...> - 2004-02-06 09:37:26
|
Hello, all,
SIP 0.3.0 release candidate 2 is out.
Generic sourcecode is available, as well as a **binary package** for GNU/Linux
systems. The changes from rc1 are:
CHANGELOG
+ new mkfftfilter function from Jocelyn.
+ build process slightly improved.
+ INCOMPATIBLE: Double-precision Images are now normalized like M4t1ab:
TYPE OF IMAGE SIZE VALUE RANGE
Grayscale images 2D 0-1
Truecolor images 3D (RGB) 0-1
Paletted images, index 2D 1-NC (Number of Colors)
Paletted images, colormap NCx3 0-1
This has many advantages. For instance, imwrite and imshow doesn't require
normalization to obscure values such as 255 or 65535.
+ Tested with animal 0.10.4, ImageMagick 5.5.7-15 and gandalf 1.1 to 1.3.2
DOWNLOAD
http://sf.net/project/showfiles.php?group_id=68055&package_id=66476&release_id=215188
I would like to release the final 0.3.0 as soon as possible, so any
of your comments are more than appreciated now.
TO JOCELYN:
I updated your functions to work with the new normalizations.
I guess they're working, but please test them yourself as well:
- GUI
- unwrap*
- poledetection
- improfile
- etc
I will wait your feedback before releasing sip-0.3.0
Did you have to use ImageMagick 5.5.7 to compile SIP?
best regards,
Ricardo.
|
|
From: DRUEL J. <joc...@li...> - 2004-02-06 06:22:24
|
Le Jeudi 05 Février 2004 17:05, vous avez écrit : > Hi all, > > I'm working translating the SIP man pages to XML format. > Now, I have a problem: few man pages (imfinfo, imread, imshow and > imwrite) have two levels inside the PARAMETERS tag. Explaining better, > my problem is to "map" this set of attributes to the dtd scilab file. > I don't have enough information (and knowledge, in most) to deal with > this. I intended to write to Ricardo, but instead I prefer to write here > to let all you know and, maybe, get some clues. Hi Vitor, Did you have a look at the formatman function in scilab 2.7 ? I did not test it, but it seems to have translation capabilities. Maybe it could help you. Jocelyn Druel -- Powered by Linux |
|
From: Ricardo F. <rf...@if...> - 2004-02-05 20:51:09
|
Hi, all, I'm almost finished with sip-0.3.0 candidate 2. Please wait I release it to continue with your testing. Vitor, Thanks for your XML translations of the help pages. However, please wait for sip-0.3.0 before proceeding. It's nice to hear that youre using PERL to translate the manual pages automatically. That will help a lot. best regards, Ricardo. -- Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil. WEB: www.rfabbri.kit.net ICQ: 208974212 "With pain and suddeness is how Real Offsprings are born. But they must result from intense pleasure and a long period of development." |
|
From: DRUEL J. <joc...@li...> - 2004-02-05 20:38:50
|
Hi, > I've tried to compile SIP myself with gcc-2.96. > GCC seems to hang for a long time, but compilation does finish. > So SIP compiles OK with this version of GCC, although > it takes a long time, probably due to a gcc-2.96 limitation. > > Ricardo. All right, I was too impatient... I did the test with gcc-2.96 on my Debian: very long but successfull (maybe Ricardo could just signal this in the INSTALL.TXT: some people (like me) might not be patient enough). Good... As I'm looking at the new sip-version, I remember that I suggested a function mkfftfilter to easily create filters (like Butterworth) for fft filtering purposes. I did not see it in the dev versions. Was it just forgotten, or is it not interesting enough (or is there a problem with it) ? I post it again with its man page. Best regards Jocelyn -- Powered by Linux |
|
From: Ricardo F. <rf...@if...> - 2004-02-05 18:52:06
|
Hi, I've tried to compile SIP myself with gcc-2.96. GCC seems to hang for a long time, but compilation does finish. So SIP compiles OK with this version of GCC, although it takes a long time, probably due to a gcc-2.96 limitation. Ricardo. On Thu, 5 Feb 2004, joc...@li... wrote: > > Compilation with gcc-2.96 did not work for me. > > >> It would be nice if you could send me again the compile > >> errors you got > >> (if you got the time).. Or those errors were the same as > >> you told me before? > > I don't know what was wrong with gcc-2.96: the computer > seemed to enter an infinte loop with cpu usage around 80%, > but I had no error message (I had to stop compilation with > CTRL+C). > If I remember well, it was when compiling mogrify, but I'll > retry it to give you more details. |
|
From: Vitor V. A. S. <vi...@cn...> - 2004-02-05 16:09:40
|
Hi all, I'm working translating the SIP man pages to XML format. Now, I have a problem: few man pages (imfinfo, imread, imshow and imwrite) have two levels inside the PARAMETERS tag. Explaining better, my problem is to "map" this set of attributes to the dtd scilab file. I don't have enough information (and knowledge, in most) to deal with this. I intended to write to Ricardo, but instead I prefer to write here to let all you know and, maybe, get some clues. I think this is not a major problem (a expect some little "bugs" after the translation), but would be nice to have it working too. To clarify, I'm writing a perl script (time to learn perl) to do the task automaticaly. Thanks a lot, Vitor Vasconcelos p.s.: Sorry about my english. I gave up to learn it. This month I start my french course. Merci, mes ami. :-) |
|
From: Ricardo F. <rf...@if...> - 2004-02-04 16:50:51
|
Hello, Jocelyn, please try to install ImageMagick 5.7.7 and tell me if it works. Installation is the traditional configure; make; make install. Then please reconfigure/recompile animal then sip. cheers, Ricardo. > Le Mardi 03 F=E9vrier 2004 12:34, vous avez =E9crit : > > On Tue, 3 Feb 2004, joc...@li... wrote: > > > Hi, > > > Personaly, I'd prefer to always stay in the 0-255 range for > > > reading and writing images. > > > > > > For intermediate results, we can still use the precision we > > > want. It was not always easy to manage reading/writing in > > > 16bits. > > > > I agree, but intermediate results are given by the Mogrify routine.. > > and it currently assumes a 0-65535 input, for quality. > > > > I'm thinking about putting a parameter to SIP on whether the user > > wants 8bit or 16bit, but leave 8bit by default.. > > OK, this seems excellent. > > > > > PS: the interferometry demo probably does not work well with > > > the new version of mogrify (because of incompatible syntax). > > > I don't think I will have the time before wednesday 4 to > > > correct it (but someone else can do it). > > > > OK, I'll wait until you can correct it. > > > > For now, I released 0.3.0-candidate1 so you can test what you can > > http://sourceforge.net/project/showfiles.php?group_id=3D68055&package_i= d=3D6647 > >6&release_id=3D214617 > > > > I'm stuck generating the binaries.. hope to get them right soon. > > > > I finally downloaded animal ans sip-devel when sourceforge had no more > problems. > > For animal, no problems compiling, thanks to the corrections Ricardo made= =2E > > But for sip, I have troubles with mogrify_int.c > First, I just tried configure, make and I got errors (I join the file > report1.txt). > My configuration is Debian Woody. Default gcc is 2.96 and imagemagick is = 5.4.4 > > So, I tried a configure CC=3Dgcc-3.0 because debian ships with a gcc vers= ion > 3.0 too. I got the same errors (file report2.txt). > > I will however try to make corrections to interferometry_demo tomorrow, b= ut I > would have liked to try it for real. > > Best regards > Jocelyn > > Ricardo. -- Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil. WEB: www.rfabbri.kit.net ICQ: 208974212 "With pain and suddeness is how Real Offsprings are born. But they must result from intense pleasure and a long period of development." |
|
From: DRUEL J. <joc...@li...> - 2004-02-04 16:01:47
|
Hi, I worked hard to make sip-0.3 work on my Debian Woody box: Everything seemed fine with configure, but I had many errors with make. So, I did it the hard way: 1) Downloaded ImgeMagick source (5.5.7) configure --with-perl=no make make install 2) I removed libmagick5-dev that ships with Debian Woody 3) I recompiled gandalf and animal (maybe it xas not necessary) 4) with SIP-0.3: ./configure CC=gcc-3.0 make make install Compilation with gcc-2.96 did not work for me. This time SIP-0.3 works fine (of course, I did not test everything). So I send the (really) corrected version of the interferometry-gui demo. Best regards Jocelyn -- Powered by Linux |
|
From: DRUEL J. <joc...@li...> - 2004-02-04 14:33:38
|
hi, I did the correction (just one line with the median filtering). It should work but I could not test. Best regards Jocelyn -- Powered by Linux |
|
From: DRUEL J. <joc...@li...> - 2004-02-03 22:10:13
|
Le Mardi 03 Février 2004 12:34, vous avez écrit : > On Tue, 3 Feb 2004, joc...@li... wrote: > > Hi, > > Personaly, I'd prefer to always stay in the 0-255 range for > > reading and writing images. > > > > For intermediate results, we can still use the precision we > > want. It was not always easy to manage reading/writing in > > 16bits. > > I agree, but intermediate results are given by the Mogrify routine.. > and it currently assumes a 0-65535 input, for quality. > > I'm thinking about putting a parameter to SIP on whether the user > wants 8bit or 16bit, but leave 8bit by default.. OK, this seems excellent. > > PS: the interferometry demo probably does not work well with > > the new version of mogrify (because of incompatible syntax). > > I don't think I will have the time before wednesday 4 to > > correct it (but someone else can do it). > > OK, I'll wait until you can correct it. > > For now, I released 0.3.0-candidate1 so you can test what you can > http://sourceforge.net/project/showfiles.php?group_id=68055&package_id=6647 >6&release_id=214617 > > I'm stuck generating the binaries.. hope to get them right soon. > I finally downloaded animal ans sip-devel when sourceforge had no more problems. For animal, no problems compiling, thanks to the corrections Ricardo made. But for sip, I have troubles with mogrify_int.c First, I just tried configure, make and I got errors (I join the file report1.txt). My configuration is Debian Woody. Default gcc is 2.96 and imagemagick is 5.4.4 So, I tried a configure CC=gcc-3.0 because debian ships with a gcc version 3.0 too. I got the same errors (file report2.txt). I will however try to make corrections to interferometry_demo tomorrow, but I would have liked to try it for real. Best regards Jocelyn -- Powered by Linux |
|
From: =?iso-8859-1?Q?<joc...@li...> - 2004-02-03 10:59:24
|
Hi,=0D=0APersonaly, I'd prefer to always stay in the 0-255 range for=0D=0A= reading and writing images.=0D=0A=0D=0AFor intermediate results, we can s= till use the precision we=0D=0Awant. It was not always easy to manage rea= ding/writing in=0D=0A16bits.=0D=0A=0D=0AOn another side, I hope "normaliz= ing" does not have the same=0D=0Amean as the function "normalize" in SIP:= I would have=0D=0Aproblems if images occupied the whole 0-255 range=0D=0A= automatically, without any user control.=0D=0A=0D=0ABest regards=0D=0AJoc= elyn=0D=0A=0D=0APS: the interferometry demo probably does not work well w= ith=0D=0Athe new version of mogrify (because of incompatible syntax).=0D=0A= I don't think I will have the time before wednesday 4 to=0D=0Acorrect it = (but someone else can do it).=0D=0A=0D=0A=0A=0A********** PROTEGEZ VOS E-= MAILS !********** =0AAvec Tiscali SuperMail, vos e-mails en toute s=E9cur= it=E9 ! =0AAnti Spam personnalisable =0AAnti Virus actualis=E9 en permane= nce =0Aet de nombreux bonus... =0APour en savoir plus, rendez-vous sur ht= tp://www.tiscali.fr/supermail/=0A |
|
From: Ricardo F. <rf...@if...> - 2004-02-03 01:18:10
|
Dear SIP codevelopers, SIP 0.3.0 is really about to come (in less than a day). I need suggestions about image normalization. ImageMagick generally stores images with 16bit integers, meaning that the images ranges from 0-65535 (2^16). This is better for precision and quality, but this isn't intuitive to many users. Shall SIP normalize the output of every ImageMagick image to 0-255? I already did this to imread. Shouldn't I also do this for the input to imwrite? This would impact precision, but is this precision really important? If not, I could even ship the SIP-0.3.0 binary together with a 8-bit ImageMagick, which is much faster. We also need someone to port SIP to windows. If you care about Windows, please help this poor Linux programmer :^) thanks for any help! Ricardo. -- Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil. WEB: www.rfabbri.kit.net ICQ: 208974212 "With pain and suddeness is how Real Offsprings are born. But they must result from intense pleasure and a long period of development." |
|
From: Ricardo F. <rf...@if...> - 2004-02-01 20:11:51
|
Greetings, another new SIP development version is released! The mogrify routine is now robust and full-working. You can now easily DRAW objects in your image using the '-draw' argument and many more Image Processing operations which previously didn't work are now available. ChangeLog: 20040201 + INCOMPATIBLE IMPROVEMENT: all arguments to mogrify are now working! But now mogrify's "argv" argument now a string vector for options with parameters. For example, suppose "image" is a scilab array: BEFORE: mogrify(image, '-rotate 60'); // WRONG NOW : mogrify(image, ['-rotate', '60']); // CORRECT This way, options like "draw" and many others are now working!! An example: mogrify(image, ['+antialias'; '-draw'; 'ellipse 50,50 30,30 0,360']); this, of course, equalizes the image and draws an ellipse, all in one command and all done in memory. If only one string is present you may type as before: mogrify(image, '-equalize'); // histogram equalization + Error treatment in mogrify is now robust and descriptive. In previous versions, mogrify didn't report errors if the parameters were wrong. + Error messages are now more descriptive in imread, imwrite, imfinfo and mogrify. + This release was tested with ImageMagick 5.5.7 and Animal 0.10.1 Download the new version at: http://siptoolbox.sf.net/devel Compilation and installation is suited for Linux, but should work correctly in other unix systems. I myself use Slackware Linux 9.1 Please report to the list any problems that you see with SIP in other distros. First please install the new AnImaL 0.10.1: http://animal.sourceforge.net You shall install the new SIP and Animal version over the older ones without problems, but if you want, you can remove the old version like this: in the unix shell, change dir to the old SIP-dev or Animal then type the following: prompt# make uninstall You may now install the new versions as usual. best regards, Ricardo. -- Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil. WEB: www.rfabbri.kit.net ICQ: 208974212 |
|
From: Ricardo F. <rf...@if...> - 2004-01-31 18:55:46
|
Hi, all, there was a download problem with the animal library at Sourceforge. I uploaded the files again and it is now woking. This was an obscure problem with Sourceforge regarding projects starting with "a".. hope you can now enjoy the new sip-dev, Ricardo. On Sat, 31 Jan 2004, joc...@li... wrote: > Hi Ricardo, > I could download sip-dev allright, but > I still can't download animal: > I send you screenshots of what I have. > > On the sourceforge status page, I found an indication of the > problem. But I think it should now be solved: > > ( 2004-01-28 13:31:59 - Download Servers, SourceForge.net > Web Site ) On 2004-01-27, an issue occurred related to > the serving of files for projects starting with the letter > 'a' from the SourceForge.net download system. This issue has > been resolved. This issue was related to an issue whereby > files posted by projects making their first release showed > to have a 0-byte size. The SourceForge.net team is working > to clear out these broken releases and will contact admins > directly via email if releases need to be reposted (this > would only affect new releases containing 0-byte files). > Further enhancements will be made by the SourceForge.net > team to improve scalability of the download server system. > > > Best regards > Jocelyn |
|
From: Ricardo F. <rf...@if...> - 2004-01-31 05:19:50
|
Hi, all, yet another new SIP development version is released! The mogrify command now works with color images. Mogrify is the richest command in SIP and provides almost all ImageMagick functionality: image rotation, scaling, filtering, segmentation, etc.. Download the new version and checkout the Changelog at: http://siptoolbox.sf.net/devel ONLY BZIP2 SOURCES NOW: tar xjvf sip-dev-2004....tar.bz2 Compilation and installation is suited for Linux, but should work correctly in other unix systems. I myself use Slackware Linux 9.1 Please report to the list any problems that you see with SIP in other distros. First please install the new AnImaL 0.10.1: http://animal.sourceforge.net You shall install the new SIP and Animal version over the older ones without problems, but if you want, you can remove the old version like this: in the unix shell, change dir to the old SIP-dev or Animal then type the following: prompt# make uninstall You may now install the new versions as usual. best regards, Ricardo. -- Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil. WEB: www.rfabbri.kit.net ICQ: 208974212 ------------------------------------------------------- The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA. http://www.eclipsecon.org/osdn _______________________________________________ SIPtoolbox-devel mailing list SIP...@li... https://lists.sourceforge.net/lists/listinfo/siptoolbox-devel |
|
From: Ricardo F. <rf...@if...> - 2004-01-31 05:00:02
|
Hi, Jocelyn, yep, SF was just havin pblms... when this sort of things happen, you might want to check the SF status page: http://sourceforge.net/docman/display_doc.php?docid=2352&group_id=1 However, the outage you said is not listed there. If you're still havin pbms pls warn me with urgence. Good news: I've been improving mogrify and it now works for color images. This means *lots* of operations are now working for these images (rotate, scale, sharpen, etc). New sip-dev is due today or tomorrow. If time permits I'll release a surprise just after it :^) cheers, Ricardo. On Fri, 30 Jan 2004, DRUEL Jocelyn wrote: > Hi Ricardo, > Thanks for the link. > > I wanted to test the new sip-dev release but I can't download the animal > library from sourceforge (I always go on error 404). I think sourceforge has > a problem (it was even hard to find a mirror to download sip-dev). > > On the mirrors, the best thing I could have was an empty directory for > /sourceforge/animal > > Do I just have to wait until sourceforge corrects these problems or is there > something on your side ? > > > > > I added the link to your manual in the SIP homepage. > > > > Ricardo. |
|
From: DRUEL J. <joc...@li...> - 2004-01-28 15:37:30
|
Hi, I started writing a guide to use SIP a few time ago. You can consult it here: http://genie-optique.chez.tiscali.fr/SIP/index.html just follow the link documentation. You can send me comments or tell me if you find errors in it. Best regards Jocelyn -- Powered by Linux |
|
From: Ricardo F. <rf...@if...> - 2004-01-27 22:18:43
|
Hi, all, a new SIP development version is out. This release brings us many fast exact euclidean distance transform algorithms, such as Olivier Cuisenaire's (1999) and Calvin Maurer's (2003). These are faster than all previous algorithms available in SIP, and are probably the best anywhere. Download the new version at: http://sourceforge.net/project/showfiles.php?group_id=68055&package_id=67061&release_id=213035 ChangeLog + Works with Animal 0.10.1 + New extremely fast euclidean distance tranform algorithms were added to bwdist: Olivier Cuisenaire's algorithm (CVIU 1999) and Calvin Maurer's algorithm (PAMI 2003). These seem to be the fastest exact methods available! There is also a nice example in the help page for "bwdist". Compilation and installation is suited for Linux, but should work correctly in other unix systems. I myself use Slackware Linux 9.1 Please report to the list any problems that you see with SIP in other distros. First please install the new AnImaL 0.10.1: http://sourceforge.net/project/showfiles.php?group_id=75126&package_id=75574&release_id=212997 You shall install the new SIP and Animal version over the older ones without problems, but if you want, you can remove the old version like this: in the unix shell, change dir to the old SIP-dev or Animal then type the following: prompt# make uninstall You may now install the new versions as usual. best regards, Ricardo. -- Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil. WEB: www.rfabbri.kit.net ICQ: 208974212 |
|
From: Ricardo F. <rf...@if...> - 2004-01-15 22:06:19
|
On Tue, 13 Jan 2004, DRUEL Jocelyn wrote: > > I hope you'll like this new version. > Hi, Jocelyn, Thanks for the new imroi. It is in the new SIP-dev-20040115 I would just like to note about versioning of the files, so contributors get to know about the methodology I use. First of all, lines like this: // $Revision: 1.4 $ $Date: 2004/01/15 21:03:37 $ are written automatically by a software suite called RCS (Revision Control System). It is much poorer than CVS, but it is much easier to use for one person alone. While we don't have time to move to CVS, we will stick with RCS. But with RCS I'm the one who commits the file into a new version. So when you send me a new file, you don't need to write the new version number in the file. It is fine to just write a Changelog line just like you did with imroi: // Changelog: // rev1.6 ellipse selection added OTOH, with CVS you will be able to commit the file directly to a central repository hosted by Sourceforge. thanks for contributing! Ricardo. -- Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil. WEB: www.rfabbri.kit.net ICQ: 208974212 |
|
From: Ricardo F. <rf...@if...> - 2004-01-15 21:55:34
|
Hi, all, a new SIP development version is out. This one brings us many important bugfixes that made a lot of SIP functions to spend a lot of memory which wasn't freed. Some of you may have noticed that some of the functions, when used inside a long loop, progressively used up a lot of the computer memory. I hope to have fixed all of these bugs. Download the new version at: https://sourceforge.net/project/showfiles.php?group_id=68055&package_id=67061&release_id=210153 ChangeLog + Fixed memleaks in various C functions (thx. to Linder). Corrected functions are: imread, bwlabel, thin, skel, bwdist, drawline, mogrify + Fixed severe memleaks in bwdist. (thx. to Linder) + Fixed Bug in imshow regarding to indexed images (thx. Marconi). As a side effect, this fixes a bug in SIP's introductory demo. + New ellipse selection in function imroi. (thx. to Jocelyn). + New ind2rgb function to convert indexed images to direct RGB hypermatrix storage. + New poledetection function to detect residues in wrapped phased images. + Many other bugfixes and minor improvements Compilation and installation is suited for Linux, but should work correctly in other unix systems. I myself use Slackware Linux 9.1 Please report to the list any problems that you see with SIP in other distros. First please install the new AnImaL 0.9.0: http://sourceforge.net/project/showfiles.php?group_id=75126&package_id=75574&release_id=209605 You shall install the new SIP and Animal version over the older ones without problems, but if you want, you can remove the old version like this: in the unix shell, change dir to the old SIP-dev or Animal then type the following: prompt# make uninstall You may now install the new versions as usual. have a nice 2004, best regards, Ricardo. -- Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil. WEB: www.rfabbri.kit.net ICQ: 208974212 |
|
From: DRUEL J. <joc...@li...> - 2004-01-13 16:59:40
|
Hi, I added an ellipse selection to imroi (which could only perform rectangular selections) and I updated the man page consequently. I hope you'll like this new version. Best regards Jocelyn -- Powered by Linux |
