Research And System Implementation Of Moving Object Tracking Based On Image

With the development of intelligent video surveillance,the image processing technologies such as target tracking are gradually connected with common application scenarios.At the same time,the new application environment has more stringent requirements for the performance of the tracking algorithm.Based on the related project,this paper implements the function of moving object detection and lock tracking in the general network monitoring system.In order to realize this function,intelligent image analysis algorithm and PTZ control technology are introduced into the system.Firstly,the moving target detection,tracking technology based on image and other related principles are analyzed.Based on actual demand,Cam Shift tracking algorithm which is simple and effective is chosen for further research.Problems of Cam Shift algorithm,such as strong environment dependence and prone to tracking drift,are existed.Therefore,this paper puts forward some improvement ideas.By introducing the Kalman filter,the local difference method and the optimal feature,the tracking effects are significantly optimized.The results of experiments show that the improved algorithm can reduce the instability in the tracking process,improving the accuracy and robustness with proper computation.On the basis of algorithm research,according to the requirements of the project,the embedded platform and PTZ camera are used to form the hardware platform.Under the premise of ensuring the stability of video acquisition,compression and transmission,the system uses multi-threaded programming to complete the program design of target detection and target tracking.Finally,the availability of the system and the portability of the tracking algorithm are proved through practical tests.Compared with the existing tracking cameras,this system has obvious advantages in technology.Therefore,it has more stable tracking effect and practical value.