Delineating Individual Trees from Lidar Data

In this research paper Marek Jackubowski and a California -based team compare raster and vector segmentation approaches with ground truth. The tree height agreement was high between the two approaches as compared to ground data (r2: 0.93–0.96).

This really is impressive, to see remote sensing come this far. For many years this was the search for the holy grail for the forestry industry. I guess once in awhile you do find it, if you work hard enough and some things go your way.

This entry was posted in remote sensing, Research. Bookmark the permalink.

4 Responses to Delineating Individual Trees from Lidar Data

  1. Bob Thomason says:

    Hi Gene,

    University of WA has been doing this for over a decade. Hans Eric Anderson, Ken Winterberger, Steve Reutebuch and Bob McGaughey have perfected this process!! Best part is, their hard labor has resulted in an awesome “public” domain software program that you can download for the UofW website. Can’t remember the name right now, but before you publish this response, let me get the name to you.

  2. Bob Thomason says:

    I hope my first post from this AM, made it through! Or the names in this post will make no sense.

    Sorry I forgot the name of the software earlier but it’s called “Fusion” was created, fine-tuned and modified by Bob! It was/is truely amazing software. It used ortho photo’s, sattilite imagery, whatever, in coordination with 15cm accuracy LiDAR, tight stuff “back then”.

    Through their affiliation with the USFS in WA and working with Ken in the AK office, A company that I used to have association with, acquired some LiDAR of the “plot” sites in different areas of AK. In these plots sites (~ 300m in size) accurate records had been kept of not only the species of tree, but details such a base (stem) diameter, height, canopy width, etc for each tree was recorded over years and then using the LiDAR they were able to calculate not only bia-mass of individual trees but potenially the cost based average analysis of an entire forest and the accuracy of this new technology LiDAR!! These “plot” sites exist throughout the US and due to the scientific nature of this project their locations are secret!!

    I feel fortunate to have been able to work with these individuals, especially since back then there was no LAS binary and Bob was working with 5-700 MB ASCII files!! The software was remarkable fast and to see his “wiggle vision” was soemthing to behold!!

  3. Pingback: Delineating trees with LiDAR | Wisconsin LiDAR Data Users Group

Leave a Reply

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.