雷达目标检测与跟踪综合处理技术的研究

发布时间：2018-02-02 19:31

本文关键词： 目标检测目标跟踪综合处理检测前跟踪　出处：《大连海事大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着21世纪航运业高速发展,雷达在水上航运中的安全效能也越来越受到人们的重视。目标检测和跟踪性能是影响雷达安全效能的两项重要因素。雷达目标检测的主要任务是从噪声、杂波干扰中发现目标,并提取目标的各种有用信息,如角度、距离、速度等。雷达跟踪的主要任务是估计和预测运动目标下一时刻的状态和确定目标的轨迹。目标检测是跟踪的前提,为了实现目标的可靠性跟踪,首先要保证对目标的准确性检测。但是对于弱目标而言,在海况复杂的杂波背景下不能有效的被检测出来,这样就无法保证对目标的可靠跟踪。因此,本文将检测与跟踪进行综合处理,以便提高弱目标的检测和跟踪性能。本文首先介绍了雷达目标检测与跟踪的基本理论与方法,包括信号检测的最佳准则,恒虚警和脉冲积累检测方法,卡尔曼滤波和机动目标跟踪的方法,接着分析了检测与跟踪两者之间的关系。然后针对检测前跟踪算法与先检测后跟踪算法的不同,本文提出了一种判别机制,判断何时启用检测前跟踪算法。接着研究了针对弱目标的动态规划检测前跟踪算法,并加入了两级门限对其进行了改进。最后为了改变转移步长,将卡尔曼滤波与动态规划进行了结合,并进行了仿真分析。本文针对基于改进的动态规划算法的检测与跟踪综合处理算法,利用Visual Studio2010平台进行了软件实现。为了加大处理速度,使用了基于CUDA的编程方法,并检验了目标检测与跟踪的效果。
[Abstract]:With the rapid development of shipping industry in 21th century. More and more attention has been paid to the safety effectiveness of radar in maritime navigation. Target detection and tracking performance are two important factors affecting radar security effectiveness. The main task of radar target detection is noise. In clutter, the target is found, and the useful information, such as angle and distance, is extracted. The main task of radar tracking is to estimate and predict the state of the moving target at the next moment and determine the track of the target. Target detection is the premise of tracking, in order to achieve the reliability of the target tracking. First of all, to ensure the accuracy of target detection. But for the weak target, can not be effectively detected in the sea complex clutter background, so it can not guarantee the reliable tracking of the target. In this paper, detection and tracking are integrated to improve the detection and tracking performance of weak targets. This paper first introduces the basic theories and methods of radar target detection and tracking, including the optimal signal detection criteria, CFAR and pulse accumulation detection methods, Kalman filtering and maneuvering target tracking methods. Then the relationship between detection and tracking is analyzed. Then, aiming at the difference between the pre-detection tracking algorithm and the pre-detection and post-tracking algorithm, a discriminant mechanism is proposed in this paper. Then the dynamic programming pre-detection tracking algorithm for weak targets is studied, and the two-level threshold is added to improve it. Finally, in order to change the transfer step size. The Kalman filter and dynamic programming are combined and simulated. In this paper, the integrated detection and tracking algorithm based on improved dynamic programming algorithm is implemented by using Visual Studio2010 platform. In order to increase the processing speed. The programming method based on CUDA is used, and the effect of target detection and tracking is verified.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN953

【参考文献】