Change Detection of Water Resources via Remote Sensing: An L-V-NSCT Approach

Abstract

Aiming at the change detection of water resources via remote sensing, the non-subsampling contour transformation method combining a log-vari model and the Stractural Similarity of Variogram (VSSIM) model, namely log-vari and VSSIM based non-subsampled contourlet transform (L-V-NSCT) approach, is proposed. Firstly, a differential image construction method based on non-subsampled contourlet transform (NSCT) texture analysis is designed to extract the low-frequency and high-frequency texture features of the objects in the images. Secondly, the texture features of rivers, lakes and other objects in the images are accurately classified. Finally, the change detection results of regions of interest are extracted and evaluated. In this experiment, the L-V-NSCT approach is compared with other methods with the results showing the effectiveness of this method. The change in Dongting Lake is also analyzed, which can be used as a reference for relevant administrative departments.