Slightly moved vehicle detection with coherent change detection on X-band high resolution SAR imagery

Satellite-borne synthetic aperture radar (SAR) is effective remote sensor for daily geo-surface measurement, even in cloudy weather or darkness. To detect moving or parked vehicle, an experimental study of change detection with two SAR images obtained different days was studied in recent literature. The targets in the report did not be controlled; therefore, sensitivity of the detection related to target position has not been treated. In addition, the conventional system only classifies the targets as changed or unchanged. In this paper, we propose and analyze an effective change detection system for monitoring slightly moved vehicle, which is about 2 cm with temporal SAR images. The proposed method uses not only coherence but also amplitude of the images. The results from experiments show the effectiveness of the proposed method.