A method to improve the accuracy of SAR image change detection by using an image enhancement method

Abstract

Accuracy in remote sensing image change detection is an important area of study. A new approach for improving the change detection accuracy of SAR remote sensing images is investigated in this paper. The research in this paper is presented in three parts. First, we proposed a new image enhancement algorithm. We combined the image enhancement algorithm based on the combination of the wavelet domain and spatial domain and the power-law. The contrast of the original image was enhanced by histogram equalization; then, the high- and low-frequency coefficients of the image were processed by wavelet fusion, and the image was sharpened by a nonsharpening mask so that the image could be sharpened from the image. Compared with other correlation enhancement contrast algorithms, we assert that the algorithm retains the high-frequency details of the image while improving its sharpness. Second, we propose a new change detection algorithm. A power rate change detection algorithm is used to improve image brightness, a logarithmic difference map is used to obtain a difference map, and an FLCM clustering algorithm is used to improve the clustering effect by suppressing noise from domain information. To improve detection accuracy, a saliency map is extracted from the difference map, the former of which is then used. The detection template obtained from the graph greatly reduces the range of change detection and effectively reduces the noise of the SAR image. The proposed power rate change detection algorithm is superior to similar algorithms. Third, our proposed enhancement algorithm is applied to our proposed change detection algorithm, which further improves the accuracy of image change detection.