Battle damage assessment with change detection of SAR images

A Battle Damage Assessment (BDA) algorithm with change detection of Synthetic Aperture Radar (SAR) images is established. Aiming at the difficulty of accurately distinguishing the attacked and non-attacked areas within a large swath SAR scene, SAR images are divided into small patches, and Scale Invariant Feature Transform (SIFT) semantic features are extracted for each of them. According to the semantic features of the two SAR images imaging before and after attack, change area is detected and evaluated in terms of battle damage assessment. The proposed method can significantly reduce the influence of the scattering intensity difference of SAR images. Moreover, damage level is judged by quantitative calculation from the change detection results, and BDA is achieved.