Research On The Target Tracking Technology Based On Radar Image Feature

Target tracking in radar system is the key issue attracting attention of the domestic and international scholars. Because the radar tracking have a wide range of applications whether in the field of military or civilian fields, such as missile tracking, safe navigation and so on. With the rapid development of information technology and the sharp rise in the performance of computer, it is possible to apply computer image processing technology to track radar target.Firstly, this paper studies the basic theory of radar target tracking. Due to the existence of measurement error of radar system itself and the inaccuracy of motion model of the target, tracking method based on data driven will lead to the instability of tracking. Therefore, the technology of image tracking is widely used in radar target tracking combined with the radar image features in order to improve the target detection and tracking performance. Secondly, the radar image tracking algorithm is going to be discussed, including three aspects:image preprocessing, target detection and feature extraction. The effective image preprocessing can reduce the computational quantity and improve the accuracy of the target tracking. Median background model, the mean background modeling and mixture Gaussian background modeling are compared in detail regarding theirs advantages and disadvantages in the background difference method, then the author analyses the inter frame difference method effect through the acquisition of the radar image. This paper will get most of the target feature extraction through median modeling method to generate the background and do difference operation with the current image. The image you get can do morphological processing. Then the target tracking is realized through matching the target being tracked and calculating the position of the target in the next frame. Then, the position of the target is prejudged by means of meanshift method, which can avoid the disadvantage of the time consuming. Aiming at the problem of target tracking failure when the mean shift method is occluded in the target area, the Kalman algorithm is introduced to predict the target position improving the robustness of the mean shift algorithm.Finally, this paper discusses radar image sequences to track through the descriptors and the meanshift tracking algorithm of radar image feature tracking. Experiments on radar image sequence tracking show that the proposed algorithm can obtain good tracking results and verify the effectiveness of the proposed method.