Research on Target Damage Assessment Based on SAR Images

SAR images not only have the characteristics of all- way and all-weather but also provide target information which is different from visible and infrared sensors. Meanwhile, evaluation of damage effect for military targets is undoubtedly crucial in information warfare. Unfortunately, exact and timely evaluation in most cases is still unavailable. The paper researches on evaluation of damage effect for battlefield targets (especially for large or medium-sized buildings, villages and town) in the SAR images. A Gauss-Markov model is taken for wavelet-based SAR image despeckling so that the following process of image segmentation is more accurate. A damage assessment method is presented based on the gray co-occurrence matrix. Experiments on SAR images obtained by homemade SAR shows the method proposed in the paper is effective.