Font Size: a A A

Research On Target Location And Tracking Algorithm Based On Binocular Camera

Posted on:2022-02-04Degree:MasterType:Thesis
Country:ChinaCandidate:Y H WuFull Text:PDF
GTID:2518306338978369Subject:Control Engineering
Abstract/Summary:PDF Full Text Request
The use of binocular camera for target positioning and tracking has the advantages of low hardware cost,high stability,accurate positioning and so on.The use of binocular camera combined with stereo matching algorithm,target detection algorithm and target ranging algorithm can achieve real-time target positioning,tracking and ranging,so it has high research value.In this paper,three aspects of target detection and tracking,target positioning,target ranging are studied.Simulation software is developed for the above three algorithms to test and demonstrate.Finally,the above three algorithms are integrated and optimized,and the actual test of the simulation verification system is completed.Among them,the target detection and tracking algorithm used is KCF target tracking algorithm.KCF target tracking algorithm has relatively fast target tracking speed,and has certain robustness to occlusion,illumination changes,etc.,which is especially suitable for target detection and tracking based on video.In the aspect of stereo matching algorithm,the advantages of BM algorithm are verified by simulation system.BM algorithm has extremely fast running speed,and the stereo matching image generated by BM algorithm can meet the requirements of this subject,so the stereo matching algorithm used in this subject is BM algorithm.In the aspect of target ranging algorithm,the principle used is "similar triangle principle",that is,the distance of the target is calculated by combining the parallax of the left and right targets of the binocular camera with the parameters of the binocular camera.Finally,the above three algorithms are integrated and optimized,and a method of dynamic selection of ROI region is proposed to greatly improve the frame rate of target tracking and location ranging.Finally,the time consumed by BM algorithm in the target tracking process can be reduced to 1/10 of the whole image matching or even lower.Through the simulation system test,the algorithm has achieved good results.Under the condition that the resolution of the binocular camera is 240 p,it can ensure the frame number of target tracking and ranging above 30 fps,and the algorithm has certain robustness.
Keywords/Search Tags:stereo matching, BM algorithm, target detection and tracking, OpenCV, target ranging algorithm
PDF Full Text Request
Related items