Tracking Before Detection Based On Dynamic Programming(DP-TBD) Algorithm

With the rapid development of radar technology,weak target detection and tracking have also been widespread concern.Compared with the traditional detection method,the pre-detection algorithm is more efficient.At low SNR target detection and tracking,it can accumulate energy through the continuous process radar data and a frame along a plurality of possible target track to achieve simultaneous detection and tracking of targets.The pre-detection tracking algorithm based on dynamic programming is studied because it is easy to implement in hardware,it is relatively small in computation and storage.In this thesis,the pre-detection tracking algorithm based on dynamic programming is studied systematically.The main contents are as follows:(1)The advantages and disadvantages of Prior to detection of the tracking algorithm before(TBD)and after detection and tracking algorithm(DBT)are analyzed..The TBD algorithm Hough transform-based 3D TBD matched filtering algorithm based on particle filter Method TBD algorithm and dynamic programming based on the four advantages and disadvantages of the proposed algorithm are analyzed.The target motion model and measurement model based on dynamic programming algorithm are built.(2)Aiming at the energy diffusion effect of the traditional pre-detection algorithm based on dynamic programming,three improved algorithms are proposed.The first improved algorithm is to introduce a new weighting factor in the traditional dynamic programming algorithm.Aiming at the shortcomings of the algorithm with low performance at low SNR,the second method which is based on the pre-detection improvement algorithm of three-frame accumulation is proposed.The algorithm improves the signal-to-noise ratio by changing the recursive accumulation of two frames to three frames recursively.In order to further improve the data correlation between frames,the third improved algorithm based on exponential smoothing is proposed.This method makes full use of the motion of the target and improves the performance of the algorithm by using the relationship between the predicted value and the value corresponding to the current frame.(3)Former multi-target detection and tracking algorithms were studied.Firstly,the extreme value method based on dynamic programming is introduced,but the disadvantage of the extreme value method is that it can't effectively detect the two targets that the target track intersects or the targets are approaching.In view of this limitation,the elimination method based on dynamic programming is proposed.The culling method first compares the intensity of the targets,detects the intensity as the target,and then removes it from the observed data.So repeat until all targets are detected.