Shadow-aided method for ground slow moving targets detection of airborne high-resolution SAR images

摘要

In the remote observation, utilizing multi-channel high-resolution SAR images to detect slow moving ground targets will be confronted with the problem of low signal to noise ratio (SNR). Fortunately, the use of the geometric relationship between the moving object and its shadow in significant size can improve the target detection performance. In this paper, based on the shadow formation model of a slow moving target, matching conditions are designed to exclude the false targets that do not satisfy them by employing the geometric relationship between the moving object and its shadow, which can reduce false alarms and improve slow ground moving target detection performance. Compared to the traditional cell average CFAR (CA-CFAR) method, simulation results show that the shadow-aided method can improve the detection performance of 2~3dB.