Automatic Segmentation of Double trunks using 3D Forest Software

[Anonymous]

I am currently working with a mobile lidar scanner which is capable of picking up around 6000 points per square meter at a walking speed of 5km per hour. Being a mobile scanner my data set is much more variable in comparison to a terrestrial lidar scanner. My current vegetation cloud data-set has a count of 11,685,616 features and terrain cloud data set has a point count of 1,870,287 features in a 30m by 30m plot.

For the processing of my data in 3D forest i used the automatic tree segmentation tool under the vegetation tab, the program manged to segment the vegetation data into trees. I then positioned this data using the position lowest points tool under the tree tab. The software gave me horizontally accurate base positions for trees that did not have double trunks.

The reason for this topic is that I encountered a problem when it came to noise points and double trunks in my data set.

The software was not able to identify trees that had double trunks. In the position by lowest points tool under the tree tab, I set the height above lowest point in cm to 130cm (Diameter Breast Height), and it still did not identify double trunk trees in my data. Do you have any suggestions to help me possibly identify the double trunk trees in my data set?

My second query is with regard to noise points, without manually adjusting my point cloud, is there any way of reducing noise within the data set, the reason for this is that the software identified more trees by automatic segmentation in comparison to how many are present in the study area. The software identified 136 trees, whilst the sample plot contains 121 trees.

I would appreciate if you have any suggestions to help me use the software more effectively for identifying trees with double trunks and reducing the noise points within my data.

Thank you in Advance.

[Janek]

Hi,
first sorry for not answering your mail. Double trunk trees can be
separated only by manual editing tool. So it depends how many double
trunks you have on plot. And for the second question. Some noise
filters are missing that true. But hopefully two filters are present
for terrain data - so you can import your data as terrain cloud, use
filter and try to import filtered cloud as base cloud. In next
version there should be this possibility for all cloud.
Jan

2018-04-16 11:13 GMT+02:00 noreply@sourceforge.net:

I am currently working with a mobile lidar scanner which is capable of
picking up around 6000 points per square meter at a walking speed of 5km per
hour. Being a mobile scanner my data set is much more variable in comparison
to a terrestrial lidar scanner. My current vegetation cloud data-set has a
count of 11,685,616 features and terrain cloud data set has a point count of
1,870,287 features in a 30m by 30m plot.

For the processing of my data in 3D forest i used the automatic tree
segmentation tool under the vegetation tab, the program manged to segment
the vegetation data into trees. I then positioned this data using the
position lowest points tool under the tree tab. The software gave me
horizontally accurate base positions for trees that did not have double
trunks.

The reason for this topic is that I encountered a problem when it came to
noise points and double trunks in my data set.

The software was not able to identify trees that had double trunks. In the
position by lowest points tool under the tree tab, I set the height above
lowest point in cm to 130cm (Diameter Breast Height), and it still did not
identify double trunk trees in my data. Do you have any suggestions to help
me possibly identify the double trunk trees in my data set?

My second query is with regard to noise points, without manually adjusting
my point cloud, is there any way of reducing noise within the data set, the
reason for this is that the software identified more trees by automatic
segmentation in comparison to how many are present in the study area. The
software identified 136 trees, whilst the sample plot contains 121 trees.

I would appreciate if you have any suggestions to help me use the software
more effectively for identifying trees with double trunks and reducing the
noise points within my data.

Thank you in Advance.

---

Automatic Segmentation of Double trunks using 3D Forest Software

---

Sent from sourceforge.net because j.trochta@gmail.com is subscribed to
https://sourceforge.net/p/forest3d/discussion/general/

To unsubscribe from further messages, a project admin can change settings at
https://sourceforge.net/p/forest3d/admin/discussion/forums. Or, if this is a
mailing list, you can unsubscribe from the mailing list.