Research On Moving Target Detection And Tracking Algorithm Based On OpenCV

The detection and tracking of moving targets belongs to the field of computer vision,which includes a variety of crossover technologies.In recent years,with the in-depth research in the field of computer vision,more and more researchers have listed target detection and tracking technology as the research content.Many new research results are widely used in the fields of uav tracking and cruise,face recognition,traffic supervision system and missile trajectory tracking,but there are still difficulties in real-time and accuracy to be solved.In this context,this paper based on OpenC V vision library for moving target detection and tracking the following aspects.Firstly,the principle of image preprocessing technology is analyzed and studied.In the aspects of image denoising,binarization and morphological changes,the advantages and disadvantages of each method were demonstrated through experiments,and the most suitable image preprocessing method was selected.In the aspect of moving target detection,the traditional detection algorithm is theoretically analyzed and the algorithm is realized.On this basis,the fusion algorithm of four-frame difference method and background average method is proposed to overcome the shortcomings of the inter-frame difference method and background difference method.The fusion algorithm USES learning rate a to update the background in real time and modifies the foreground by combining morphological processing,which can effectively deal with the slow change of background.In the aspect of moving target tracking,the advantages and disadvantages of Meanshift,Camshift and KCF filtering algorithm are analyzed.Aiming at the limitations of Camshift algorithm,Kalman filtering algorithm and Camshift algorithm are proposed to combine for tracking.On the basis of the foreground image in the detection stage,Camshift algorithm is used to accurately track the moving target.When the target is blocked by the background object,Kalman filter makes motion prediction for the target to prevent the loss of the target.When the background is still or slowly changing,the improved detection algorithm combining the four-frame difference and the average background method can effectively detect the moving target.On this basis,the fusion algorithm combining Kalman filtering algorithm and Camshift algorithm can accurately track the target in video.Finally,the detection and tracking experiments of video are carried out to verify the effectiveness of the algorithm.