Videosar Moving Target Detection from Geometric Distortion Image Sequence

Image geometric distortion in sequential image frames, which is caused by the movement of radar platform and its distance to scene center in VideoSAR, can bring great chanllenges to the moving target detection. In this paper, a VideoSAR moving target detection method is proposed, which can effectively extract moving target shadow from geometric distortion image sequence. Firstly, we utilize the speeded up robust features(SURF) method to register the image sequence in order to diminish the effect of the geometric distortion in image sequence on the correlation of the interframes. Secondly, sparse component, containing moving target and residual noise, is seperated from the low-rank stationary background by using the robust principal component analysis (RPCA). Thirdly, the residual noise in sparse component is further suppressed by performing the median filtering method. Simulation results based on the VideoSAR experimental data released by the Sandia National Laboratories (SNL) illustrate the effectiveness of the proposed method.