Research And Implementation Of Target Tracking System In Recording And Broadcasting System

With the rapid development of national economy and the gradual enhancement of scientific and technological strength,intelligent systems such as video surveillancebased on intelligent video analysis technology playan increasingly important role in the fields of distance education,security and other fields.With the advent of a variety of teaching methods such as high-quality classes,micro-grid teaching,educators who demand the intelligent recording and broadcasting system are more and more.In the promotion of technology of computer vision,the technology of target detection and tracking used in intelligent video analysis is also increasingly mature,which further improves intelligence of the recording and broadcasting system.The system of target tracking undoubtedly plays an important role in the intelligent research and development of recording and broadcasting systems.It uses the technology of video image analysis under multiple cameras to control the recording and broadcasting system to automatically perform screen switching and video recording tasks.Aiming at target tracking system,the main research contents carried out in this paper are as follows:Firstly,the technology of moving target detection in target tracking system is studied in this paper.For the system is used in indoor,several object detection methods based on static background are analyzed.By comparing the methods,the method of background subtraction combined with the Mixturn Gauss background model is finally determined as the method of target detection in this paper.Aiming at the time-consuming problem of traditional mixture Gauss background model,the method of intercepting the small block image in the large image is used to improve the processing speed greatly.Because the mixture Gauss background model will update the stationary targets,this paper introduces a feedback mechanism,which makes the static target background isn't updated.Under the improvement,the stationary target can't be updated as the background.In view of the phenomenon of "ghosting" appearing in the mixture Gauss background model,the learning rate of step change is introduced to make the “ghosting” quickly disappear,which makes it not affect the effect of target detection.Secondly,this paper studies thetechnology of moving target tracking in the target tracking system.In view of the high computational complexity of the traditional target tracking algorithm,then depending on the usage scenarios of indoor,this paper proposes a new target tracking method based on the inter-process communication,we realize the detection and tracking of targets in two processes,the target detection process extracts the foreground and feature of targets,and passes the target position and direction of rotation of PTZ instructions to the target tracking processthrough the shared memory,the target tracking process drives PTZ to achieve continuous tracking.The system experiments show that the method can achieve smooth and stable target tracking.Finally,based on the practical indoor application of the target tracking system,after completing the task of writing the code of the target detection andtracking program,the hardware resources of the embedded platform based on Hi3531 A and the software environment in Linux server are studied,then the teacher detection and tracking application is downloaded to the platform.Combined with the switch and control strategies,the system-level algorithm is tested.The experiment shows that the system can not onlyaccurately track the location of the teacher on the platform,but also achieve the tracking intelligence Switch.