Moving Target Tracking Algorithm And Application Research

Firstly,this paper introduce the development of target recognition and tracking methods at home and abroad,and correlation algorithms is introduced simply.In the aspect of target recognition and tracking,particle filter tracking algorithm and Camshift tracking algorithm are introduced and analyzed in this paper,and several comparative tracking experiments are carried out to illustrate the advantages and disadvantages of algorithm.The experimental results show that the particle filter tracking algorithm and the Camshift algorithm can effectively track the target when the color difference between the target and the background is large,but when the color difference between target and background is small,The accuracy of target tracking is greatly reduced,and even lose the target.In order to improve the stability and accuracy of the above two tracking algorithms in complex background,two improved algorithms are proposed in this paper.The first algorithm is particle filter tracking method based on the saliency histogram.The saliency weights of hues in the histogram are determined by comparing the distribution of the hues in the target and the background.In this way,the saliency histogram can be established.The saliency histogram can restrain the disturbance from the background to the target by strengthening the recognition role of the hues existing only in the target.Thus,the accuracy of the target location can be improved.The second algorithm is improved CamShift tracking method based on the edge suppression.The gray value of the target edge in the back projection image is suppressed by the weighted method according to the position and size in the previous frame.The suppressing edge can effectively separate the target and the background,and weaken the trend of the centroid outward shift.Simulation results show that the method proposed in this paper can can improve the accuracy and stability of target tracking in the complex background with the low computation cost,and the average tracking computation time can well satisfy the real-time requirement.At last,two improved tracking algorithms is applied to the tracking of intelligent car.Experiments show that the improved algorithm can obtain better tracking performance in practical applications.