Mobile Robot Motion Detection And Tracking Under Dynamic Background

Robot technology is developing rapidly in recent years,In all aspects of work and life,robot technology has a wide range of applications.So intelligent monitoring of robots is particularly important.In the case of this paper,Based on the insight into static background motion target detection,we launched under the dynamic background of target detection and tracking,Using algorithm of thed etection and tracking algorithm,this paper achieve a intelligent monitor systemo f mobile robot.Firstly,it studied the static background of mobile robot detection method,based on the inter frame difference method the background difference method of their advantages and disadvantages are compared.Through comparison,we choose the ViBe method to finish the detection of the static background of mobile robot.Secondly,we begin our study of dynamic background of mobile robot detection on the basis of the static background of mobile robot detection.Dynamic background of mobile robot detection mainly global motion compensation,so as to solve the that the background of the problem of the large amplitude motion due to camera motion,by compensating the background becomes static,and then detect mobile robot using the inter-frame difference method.Again,identify the mobile robot in the scene in a plurality of moving objects.Identification of a mobile robot mainly adopts the support vector machine(SVM)based on LBP feature and color histogram method.LBP feature has good robustness to illumination change,and computing speed is fast.Combined with color histogram,it is a better choice to describe the target characteristics.After identifying,we want to track the mobile robot.This paper use Kalman filter combining with KCF methond to track the target,improve the tracking accuracy and the adaptability of the tracking frame.Finally,the experimental set up video surveillance platform,using QT interface development tools,developed a simple,strong control of the monitoring interface.The real-time performance and robustness of the image detection,recognition and tracking of the mobile robot are verified by experiments.