Time-frequency analysis based image processing for maneuvering target detection in HF OTH radar

In this paper an algorithm for maneuvering target detection in HF over-the-horizon radar is presented, which is based on the interpretation of the time-frequency map as an image. In the algorithm, radar signal is processed in joint time-frequency domain with short-time Fourier transform. Low-threshold CFAR combined with local thresholding processing transform the time-frequency map into multi-level image. Then neighboring search based region growing technique is developed to extract target ridge among false alarms due to transient events. Real data results show that the presented multi-stage scheme can achieve accurate instantaneous Doppler estimation and good detection performance.