Radar target detection in sea clutter environment based on convolutional networks

After the radar echo signal is preprocessed, a random data feature distribution (DFD) exists for the target in the nearshore clutter environment. This distribution can be used as a detection feature to improve the detection performance. This paper proposes a fast detection scheme to detect the target using the selected DFD matrix. Using this information of DFD, a convolutional network (CNN) is designed to learn the DFD. The experimental results show that the proposed target detection method based on data feature distribution has better detection performance compared with the traditional target detection (TTD) methods and fully convolutional network (FCN) target detection methods, and the small-model convolutional network has more efficient real-time performance with fewer parameters in comparison to the FCN.