Multi-Target Detection Using 2-D Distributed Cell-Averaging CFAR in High Resolution SAR Images

Inhalt:
A local cell-averaging fast CFAR scheme is proposed for multi-target detection. The proposed algorithm is based on global censoring scheme to avoid multitarget situation in reference cell. Fast CFAR scheme is employed to reduce computational load. Local average in the vicinity of test cell can remove small sized false alarms so that additional binary clustering can be rapidly perfomed for the reduced false alarms. The test results using real SAR images show good performance in multitarget situation and heterogeneous environment.