TERRAIN CATEGORIZATION CAPABILITIES OF LIDAR SYSTEMS OVER DENSELY VEGETATED AREA

Thumbnail Image
Authors
Sonarch, Chedpong
Subjects
LiDAR
terrain classification
LiDAR system
LiDAR waveform
Advisors
Olsen, Richard C.
Metcalf, Jeremy P.
Date of Issue
2019-12
Date
Publisher
Monterey, CA; Naval Postgraduate School
Language
Abstract
Light Detection and Ranging (LiDAR) technology has various useful applications, such as in surveying tasks. This study continues research previously conducted at the Naval Postgraduate School by Andrew S. Davis, and documented in his thesis, “Forestry Identification with LiDAR Waveform and Point Clouds.” The present study, which aims to evaluate the classification capability of LiDAR systems over various tree species in a particular area of interest, collected sample data over Point Lobos State Park, California. The data set was separated into two categories, aerial platform and ground survey. The aerial platform consisted of an Optech Titan system and Airborne Hydrography AB Chiroptera system (AHAB). Analysis was performed by comparing the results from the ENVI software classifier and the actual location of tree species from ground surveying. The study also extracted the features of waveforms from each tree species and used these features to distinguish them from among other samples of tree species and their surrounding environment, such as roads and trails. The classifications were done by the classifier tools provided in ENVI (K-means, Spectral Angle Mapper, and Support Vector Machine). The results showed that waveform data can accurately distinguish class samples. The analysis also pointed out that the most common error occurred when classes had a narrow gap in data values and shared similar pulse shapes.
Type
Thesis
Description
URI
https://hdl.handle.net/10945/64073
Series/Report No
Department
Physics (PH)
Organization
Identifiers
NPS Report Number
Sponsors
Funder
Format
Citation
Distribution Statement
Approved for public release; distribution is unlimited.
Rights
Copyright is reserved by the copyright owner.
Collections
Theses