TERRAIN CATEGORIZATION CAPABILITIES OF LIDAR SYSTEMS OVER DENSELY VEGETATED AREA

Chedpong Sonarch, LTJG, Thailand Navy

LiDAR system is a technology which can fulfil various needs. The surveying tasks are one of the applications for which LiDAR can be applied. This study intends to continue working on the previous study of Andrew S. Davis, who conducted the study of "Forestry Identification with LiDAR Waveform and Point Clouds." The study aims to evaluate classification capability of LiDAR system over various tree species. The collection of sample data was collected over Point Lobos State Park, California. Dataset were separated into two categories, aerial platform and ground surveying. The aerial platform obtains from Optech Titan system and Airborne Hydrography AB Chiroptera system (AHAB). The analysis was performed by comparing the results from ENVI software classifier and actual location of tree species from ground surveying. The study also extracted the feature of waveform of each tree species which are used to distinguish them among samples of tree species and their surrounding environment; such as road and trail. The classifications were done by the classifier tools provided on ENVI (K-means, SAM, and SVM). The results show that the waveform data can achieve distinguish class samples. The analysis can point out the error most occurred between the closed class.

Point of Contact:

Advisor: Dr. Richard C. Olsen

Added:

Jan 12, 2020

Download Video
Video Tags Assigned:
Related Assets
Asset Publisher