Research On The Comprehensive Processing For Radar Target Detection And Tracking

With the rapid development of shipping industry in the21st century, people pay more attention to the radar security effectiveness for water maritime security. Target detection and tracking performances are two important factors in safety effectiveness of radar. The main task of radar target detection is to find target from noise and clutter interference, and extract all kinds of useful information, such as angle, distance, speed and so on. Radar tracking is to estimate and forecast moving target’s state of the next moment and confirm target’s trajectory. Target detection is the premise of target tracking. In order to achieve the reliability of tracking, the accuracy of target detection should be ensured. But for the weak target, under the background of complex sea clutter, the available detection is more difficult and the reliable target tracking is not guaranteed. Therefore, the comprehensive treatment of detection and tracking is raised by the paper to improve the detection and tracking performance for weak targets.Firstly, this paper introduces the basic theory and method of radar target detection and radar tracking, including optimal criterion of signal detection, constant false alarm and pulse accumulation detection method, kalman filter and maneuvering target tracking methods, then analyzes the relationship between detection and tracking.In view of the track-before-detect and the track-after-detect algorithm, a decision mechanism is proposed to determine when to enable track-before-detect algorithm. Then a improved track-before-detect algorithm is studied based on dynamic programming for weak target, and it is improved by adding two levels of threshold. Finally in order to change shift step length, kalman filtering is combined with dynamic programming and simulation is made.Finally, comprehensive detection and tracking processes based on the improved dynamic programming algorithm is carried on the software implementation based on Visual Studio2010platform, by using the method of CUDA in order to increase processing speed, and the effect of target detection and tracking on computer is tested.