Change detection in multi-temporal remote sensing images based on the wavelet-domain immune clonal optimazition

In order to reduce the impact of noises on detection accuracy in change detection for remote sensing images, a novel change detection method is proposed based on the immune clone algorithm and wavelet transform. Firstly, the multi-scale and low-pass smoothing characteristics of wavelet transform are utilized to construct multi-layered difference images. Secondly the time domain deviations caused by operation of zero insertion and image convolution in wavelet transform are corrected by the immune clonal algorithm that the initial segmentation results obtained using the Rayleigh-Gauss model are matched by the secondary linear interpolation operation. Finally, the change detection is accomplished by image fusion. Simulation results show that the algorithm can not only reduce the image noises, but also suppress the image deviations caused by wavelet transform effectively. The accuracy of change detection is improved significantly.