Detail-preserving change detection from amplitude SAR images

This paper presents a modified version of the information-theoretic feature in [1] for change detection in multitemporal synthetic aperture radar (SAR) imagery. The proposed method is capable of capturing small-area structural changes between the two images, by maintaining at the same time the capability of rejecting statistical changes due to speckle patterns or co-registration inaccuracies. This improvement is obtained by bootstrapping the original procedure by means of an adaptive preliminary selection of potentially changed pixels driven by the logarithm of the pixel ratio. Experimental results have been carried out on true SAR images acquired by the COSMO-SkyMed constellation.