Fusion of Difference Images for Change Detection

The Land use/ Land cover change in urban areas and the difference of the earth surface after the flood can be detected from remote sensing images by performing image differencing algorithms. Although many algorithms were proposed to generate difference images, the results are inconsistent. In order to integrate the merits of difference algorithms, fusion techniques are used to merge multiple difference images. The image fusion algorithms applied here are based on Principal Component Analysis and Discrete Wavelet Transform. Principal Component Analysis is the unsupervised technique, the change is guaranteed to be preserved in the major component images. In Wavelet based method, image fusion is performed at the pixel level and the details from source images can be reserved at various scales. The algorithms are implemented on the satellite images and results are presented.