Research on change detection based on wavelet transformation

摘要

Based on comparative analysis of precision and direction selection characteristic of image frequency feature in the course of wavelet transform, combined with important influence factors of sub-space estimatation during independent component analysis — decomposing of image block, the author puts forward two schemes on change detection of remote sensing image by use of wavelet transform, and analyzes application scope and realization difference deeply, and interprets improves results of image change detection by use of two schemes on the basis of theory. Simulation results show that application scope in second scheme is broader than that in first scheme because of parallel algorithm and robustness for massive remote sensing images.