Multi-Channel SAR Moving Target Detection by Integrating STAP and Faster R-CNN

The combination of synthetic aperture radar technology and ground moving target indication technology has important application value in military reconnaissance and traffic management. Compared with single-channel SAR-GMTI system, multi-channel system has better overall performance in moving target detection and parameter estimation. However, traditional methods still have some limitations in the type of target motion and the accuracy of detection. This paper proposes a detection method by combining space-time adaptive processing and deep neural network. Firstly, the multi-channel SAR echo signal including moving target and ground clutter are modeled and analyzed, and STAP is used to suppress clutter. Then the moving target is detected and recognized by applying Faster R-CNN network on the clutter-suppressed image. The experimental results show that the proposed method has high detection probability, which verifies the feasibility and effectiveness of deep learning algorithm in multi-channel SAR moving target detection.