Automatic Target Detection Based on Constant False Alarm Rate in High Resolution SAR Images

In this letter we develop a target detection method using high-resolution radar images based on constant false alarm rate (CFAR) algorithm. CA-CFAR، the specific CFAR algorithm utilized in this paper، is a CFAR method with high performance and easy implementation used in many studies providing good results. This adaptive algorithm assigns a pixel as a target pixel by comparing the pixel intensity with a threshold calculated from surrounding pixels of the pixel under test. Experimental results show that the algorithm provides best results with lowest false alarm using the “span” image that contains more polarimetric information as compared to the single channel data. RADARSAT-2 polarimetric SAR data acquired over the San Francisco city has been chosen to assess the capability of the proposed approach. The coherency matrix data is extracted from this dataset and the proposed approach is performed on single band images corresponding to different polarimetric channel and the span image as a compound of all the single polarization images.