Title :
Extracting fractal texture parameters from Radarsat images for landuse investigation in Zhaoqing test site of south China
Author :
Xiangtao, Fan ; Yun, Shao
Author_Institution :
Lab. of Remote Sensing Inf. Sci., Acad. Sinica, Beijing, China
Abstract :
The fractal parameters are extracted from 5 scenes multi-temporal fine mode Radarsat images using the method based on superficial area-volume relation. The results show that the fractal dimension reveals the surface roughness information of the image plane and the constant value exhibits the edge information clearly. Putting the fractal parameters such as texture parameters into a neural net classifier for landuse investigation in Zhaoqing test site, it indicates that the average accuracy of classification of 11 targets in this area is improved from 86.66% to 96.19%
Keywords :
fractals; geophysical signal processing; geophysical techniques; image classification; image texture; neural nets; radar imaging; remote sensing by radar; spaceborne radar; synthetic aperture radar; terrain mapping; China; Radarsat image; SAR; Zhaoqing test site; fractal dimension; fractal parameters; fractal texture parameters; geophysical measurement technique; image classification; image texture; land surface; landuse; neural net; radar remote sensing; spaceborne radar; surface roughness; synthetic aperture radar; terrain mapping; Backscatter; Data mining; Equations; Fractals; Layout; Least squares methods; Remote sensing; Rough surfaces; Surface roughness; Testing;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2000. Proceedings. IGARSS 2000. IEEE 2000 International
Conference_Location :
Honolulu, HI
Print_ISBN :
0-7803-6359-0
DOI :
10.1109/IGARSS.2000.860517
