一种改进积累与判决的动态规划检测前跟踪算法

A Dynamic Programming Track Before Detection Algorithm Based on Improved Accumulation and Sentence

由于传统的动态规划检测前跟踪算法(DP-TBD)采用多帧积累提高信噪比,所得到的目标函数在目标轨迹末端形成的峰值并不是特别尖锐,会在对后续的门限判决回溯目标轨迹时出现虚假目标。首先简单介绍了DP-TBD算法,并提出一种改进的算法,通过改进相邻2帧迭代积累为相邻3帧进一步提高信噪比。在门限判决时,不是直接对目标函数进行门限判决,而是提取目标函数局部最大的极大值,然后通过对这些值进行门限判决,从而可以减少虚假目标个数进而提高检查性能,最后通过单个运动目标和两相交运动目标仿真验证了该算法改进的可行性和有效性。

As traditional dynamic programming track before detection （DP-TBD） algorithm uses multi-frame accumulation to improve the signal-to-noise ratio （SNR）, the peak value of the target function formed at the end of the target trajectory is not particularly sharp,false targets will appear in the subsequent threshold decision backtracking to target trajectory. This paper firstly introduces the DP-TBD algorithm, and proposes an improved algorithm, which further improves the SNR through improving adjacent two frames iteration accumulation to adjacent three frames. In the threshold decision,it is not directly to perform threshold decision to the objective function,but the maximum value of the local maximum of the objective function is extracted,then the threshold decision is performed to the values, thus the false target number can be reduced and the check performance is improved. Finally, the feasibility and effectiveness of the proposed algorithm are verified through the simulation of single moving target and two intersecting moving targets.