Remote-Sensing Image Change Detection With Fusion of Multiple Wavelet Kernels

Quality of the difference image determines the potentials of the change detection algorithms. The subtraction operation and ratio operation are two commonly used tools for producing the difference image. However, the complementary information existing in the two difference images, the subtraction image and the ratio image, has found limited applications in real tasks by now. Therefore, a method which utilizes multiple wavelet kernels for fusing the complementary information of the two difference images is proposed in this paper for remote-sensing image change detection. First, the complementary information of the two difference images is analyzed. That is, the subtraction operation highlights the changed areas and the ratio operation suppresses the disturbance of the complex background. Then, for each difference image, wavelet kernels at multiple scales are computed followed by a reliable scale selection scheme based on correlation coefficients. After that, the two difference images' wavelet kernels at reliable scales are fused under the supervision of an initial change detection result. The obtained kernel, the MWF kernel, is of good homogeneity and smoothness on the changed areas as well as great suppression of the complex background's disturbance, since it takes into account the complementary information of the two difference images. The principal component analysis (PCA) and k-means clustering act on the subtraction image to produce the initial change detection result. Finally, the fused kernel is inputted into a classification algorithm based on the minimum Euclidean distance in the kernel space to get the final change detection result. Experiments demonstrate the effectiveness of the proposed method and illustrate that it possesses both strong disturbance immunity and good homogeneity of changed areas for remote-sensing image change detection.