HI Joseph/All,

 

This is in an interesting topic because it also represents the cost/time investment that you will need to count trees.

 

Lidar Analyst (below) is a great program (which I have used), and is fairly accurate. The downside… it costs $6k per license and is an Arcgis extension, not a standalone program. That being said you can extract pretty much anything from the point cloud (topography, buildings, classified landscape, tree points) with built in modules and can project it on a built in 3d viewer.

 

The reverse watershed method works fairly well, but takes some time to go through the workflow, and you still will get about 70-80% accuracy depending on the size of the point cloud. I provided a good tutorial that describes the methodology below.

 

http://web.pdx.edu/~jduh/courses/geog493f12/Projects/SwamerHouser.pdf

 

Some other tools that are useful which may save you time and are not $6k, are using FUSION which is a free LiDAR viewer and extraction program from the USFW. The link is below. The learning curve is steep as it is command line driven, but it works really well.

 

http://www.fs.fed.us/eng/rsac/lidar_training/

 

Another low cost option is the LAStools toolbox you can download for free. It contains a suite of tools for use in Arcgis that streamline modelling and cut the time for developing your final model you will use for counting tree points. Link can be found here.

 

http://www.cs.unc.edu/~isenburg/lastools/

 

 

All these options are very cool and useful, however, I think all of them still have a bit to go as far as accuracy is concerned. I believe these are all great for a first pass and hopefully in the future they will get even better!

 

 

Good Morning all!

The following question was posted on one of my LinkedIn pages: “How to do the Tree counting using LiDAR Point cloud?

Tristan

Tristan Stuckenberg

GIS Specialist at Esri South Africa

Hi Mahipal,

There are several software packages that claim to be able to do this, how well depends on the density of laser return and how clearly individual trees are captured. As a general rule of thumb if you can't clearly distinguish individual trees in the point cloud you’re going to have a hard time. I know TerraScan and ENVI LiDAR both have modules to identify and subsequently count trees. An alternative is to use FUSION, which is free, if you lack either of the aforementioned software packages. You can download it from MailScanner has detected a possible fraud attempt from "www.linkedin.com" claiming to be http://forsys.cfr.washington.edu/ and you will have to create a Canopy Height Model (CHM) and then use the CanopyMaxima script in DOS to count trees. To optimise results you will likely have to modify the moving window parameters considerably.

An alternative approach is to use watershed segmentation on a CHM but I haven't personally had much luck with this approach. Good luck and let me know how it goes.

Jonathan Gale

Solution Engineer and Part-time Lecturer, College of Professional Studies at Northeastern University

Try LIDAR Analyst. It will extract and attribute approx 10,000 tree points per minute. You can download a demo copy for free to run through the tutorials and see if it is what you are looking for. You can then request an evaluation license to test the software on your own data. During your evaluation period you can work with the product support team to help you optimize your results and develop proof of concept. Otherwise there used to be a good research article out there by Yunsheng Wang that shows you how to do this with ArcGIS and Fusion. It's a tough nut to crack but I've helped people do it for quite a while now - before LIDAR was cool ;

 

 

Joseph Young

Executive Director

Maine GeoLibrary

SHS 145

51 Commerce Drive

Augusta, Maine 04333- 145

VP 207-624-2664

FAX 207-287-1131

Subscribe to our List Serve learn more about the GeoLibrary

http://lists.maine.edu/cgi/wa?A0=geolibrary-l

 

----------------------------------------------------------------------- This list Maine GeoNews (GEOLIBRARY-L) is an unmoderated discussion list for all Maine GIS Users. If you no longer wish to receive e-mail from this list, you can remove your name and email address yourself at the this web address:

http://lists.maine.edu/cgi/wa?SUBED1=geolibrary-l&A=1

Or,you may also request that your name be removed by sending email to:

[log in to unmask]