| Moving target detection technology is a popular research direction in the field of computer vision.Its main purpose is to detect moving targets as foregrounds and extract them from video sequences.The surrounding environment of the moving targets is used as the background.It is separated from the moving targets.It has low cost,It can replace manual labor,all-weather ser-vice,etc.,and has a very broad market prospect in the application field of intelligent monitoring systems.The function of video stabilization technology is mainly to solve the problem of video image shifting due to camera shake.Video stabilization technology analyzes the global video information to find out the jitter frames in the video sequence,and combines the global infor-mation to repair the jitter frames,Can better provide a video sequence that removes interfer-ence for moving target detection technology.Moving target detection technology and video im-age stabilization technology are the current research hotspots in the field of computer vision.They are widely used in the fields of vehicle assisted driving,robots,and smart transportation,and have important economic value and research significance.In actual application of moving target detection technology,because the scene and camera equipment are often interfered by outside the system,for example,the vibration generated by the start of large equipment such as cranes in the production workshop causes the camera to shake,which causes interference and affects the effect of moving target detection.Therefore,according to the dynamic background,this article adopts video image stabilization technology to realize the detection and alarm of personnel entering the designated production area of the production workshop.Vibe(Visual background extractor)algorithm is a simple and fast back-ground modeling and fast and effective foreground judgment method.However,the Vibe mov-ing target detection algorithm will cause false detection of moving targets when the background is dynamically shifted.In response to this problem,this article proposes to introduce ORB(Ori-ented FAST and Rotated BRIEF)feature point matching improved Vibe algorithm to achieve moving target detection in a dynamic background.Extract feature points according to ORB al-gorithm,use RANSAC(Random Sample Consensus)method to calculate transformation matrix parameters,obtain global motion compensation image through perspective transformation,and model of Vibe algorithm The update strategy is combined to finally eliminate the interfer-ence information in the dynamic background.This paper uses OpenCV3.4.6 combined with C/C++standard programming language to implement the detection algorithm in the Visual Studio 2015 software environment,and uses the sample video to test the algorithm.Simulation and experimental results show that:com-pared with the Vibe algorithm moving target detection method,the algorithm proposed in this paper can remove a large number of foreground misjudgment points,improve the accuracy of moving target detection,and retain the integrity of the moving target. |
PDF Full Text Request