|
|
|
Urban Feature Characterization using High-Resolution Satellite Imagery: Texture Analysis Approach
Kiwon Lee
Dept. of Information
Systems
Hansung University
Seoul, Korea, 136-792
kilee@hansung.ac.kr
So Hee Jeon
Dept. of Earth Science
Education
Seoul National University
Seoul, Korea, 151-742
Jsh1118@snu.ac.kr
Byung-Doo Kwon
Dept. of Earth Science
Education
Seoul National University
Seoul, Korea, 151-742
bdkwon@snu.ac.kr
Abstract
As the multiple types of remotely sensed imageries have been widely used, analysis methods in the detailed level are needed in viewpoint of the image analysis and interpretation. In this study, texture image analysis scheme is applied in order to investigate urban feature characterization based on high-resolution imageries. As for texture analysis in the spatial domain based on kernel operation, second order texture based on GLCM (Grey Level Co-occurrence Matrix) and GLDV (Grey Level Difference Vector) is mainly dealt with in this study. We first developed Window-based application program for texture measure and image creation of GLCM or GLDV with multiple functions for selection of several texture parameter. Especially, in this program, texture image by omni-direction can be produced to reduce complicated directionality problem in the interpretation of texture image. By using this implementation, texture images in an urban environment with complex features are analyzed and discussed. As well, HIS-based data fusion using texture images composed of different kernel size and parameter provides distinguishable features in the visual interpretation, though further works are still left.
|
|
|