Superpixel based unsupervised change detection of manmade targets on satellite images

In this paper, a novel solution to the problem of change detection in bitemporal satellite images is presented. The approach can be described as 1) Preprocessing, in which both images´ luminance levels are approximated to the same distribution via the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm and noise is filtered out by the use of a Bilateral Noise Reduction (BNR) filter; 2) Extraction of a joint superpixel map of both images using a novel method; 3) Calculation of a variety of metrics at the superpixel level that are directed towards finding man-made changes; 4) Calculating a final change mask from a final difference image via thresholding. The proposed method is tested on bitemporal images of a number of urban regions.