Download (PDF 1.35 MB): https://research.fs.usda.gov/download/treesearch/15916.pdf
Abstract
A highly-automated procedure for remotely estimating individual tree location, crown diameter, species class, and height has been developed. This procedure will involve the use of a multimodal airborne sensing system that consists of a digital frame camera, a scanning laser rangefinder, and a position and orientation measurement system. Data from the multimodal sensing system will be processed by a model-based procedure for the recognition of individual tree crowns in aerial images. Both the multimodal sensing and tree crown recognition components have been separately tested, and data to test their integration have recently been acquired as part of an in-progress experiment. The tree crown recognition procedure has been extended to use data from the scanning laser rangefinder to increase the reliability of individual tree crown isolation and to estimate tree height. Depending on the results of the experiment, the integrated system may be useful in many forest inventory activities, delivering data at a low cost relative to its value.
Parent Publication
Citation
Pollock, Richard. 2000. Development of a highly automated system for the remote evaluation of individual tree parameters. In: Hansen, Mark; Burk, Tom, eds. Integrated tools for natural resources inventories in the 21st century. Gen. Tech. Rep. NC-212. St. Paul, MN: U.S. Dept. of Agriculture, Forest Service, North Central Forest Experiment Station. 638-645.
