Change Detection for High Resolution Remote Sensing Image Based on Co-saliency Strategy

Abstract

Change detection for remote sensing image is of great significance to a diverse range of applications. From the point of object-based method, this paper provides a change detection algorithm based on co-saliency strategy for multitemporal high resolution images. Firstly, the final difference image fused by difference feature and log difference feature, is generated, and feature image including spatial and contextual information is obtained by Gabor wavelet transform. Secondly, co-saliency strategy is performed via cluster-based method, combining the final difference image with feature difference image at each temporal data, and highlighting the common regions as the changed directly. Finally, actual change map is extracted by fuzzy local information C-means clustering algorithm (FLICM) and decision-voted method. The experiments show the method proposed in this paper has a superior performance in change detection for high resolution remote sensing images.