학술논문
Analysis and error assessment on the use of segmentation for estimating forest structural characteristics from lidar and radar
Document Type
Conference
Author
Source
2012 IEEE International Geoscience and Remote Sensing Symposium Geoscience and Remote Sensing Symposium (IGARSS), 2012 IEEE International. :5337-5339 Jul, 2012
Subject
Language
ISSN
2153-6996
2153-7003
2153-7003
Abstract
This paper investigates the ability of radar image segmentation to produce meaningful, structurally homogenous objects with respect to lidar-derived forest metrics. A comparative approach is taken to determine if radar-derived segments perform better in this respect than arbitrary, square segments or landcover-derived segments. It is found that segmentation of UAVSAR co- and cross-polarization backscatter magnitudes results in increased lidar homogeneity on the segment level relative to the arbitrary and landcover segmentations.