Design And Implementation Of Intrusion Detection System For Video Surveillance Area Based On MFC And OpenCV

As an important branch of computer,intelligent surveillance video has attracted much attention from the field of study.Video surveillance can make the corresponding response by analyzing,processing and researching the captured video files mainly using the computer vision and algorithms of image processing.For example: the monitoring system can automatically identify the invasion of moving objects;it can send real-time alarm information in the appropriate way when the monitoring area is not normal,so can provide technical support for short-term resolution of emergencies.The paper makes a detailed study about the problem of detection and processing of intrusion target in intelligent monitoring system.Video area intrusion detection system is divided into management module and image processing module.The video management module can be divided into video capture module,video storage module and video display module.The video capture module mainly realizes the system to capture the video information by reading the camera selected by the user.The main function of the video storage module is that the system can store the image frame processed by the image analysis module into the specified folder.The main function of the video display module is that the system can display video source file selected by users on the system interface,and users can view the real-time monitoring video images and the continuous image state after processed images.The image processing module is divided into image analysis module and data storage module.The main function of the image analysis module is that the system can analyze the captured video image by image processing algorithm.The main function of the data storage module is that the system can store the valid information analyzed by the image analysis module in real time in the folder.In addition,the design and implementation of system can be considered as the design of interface and the implementation of algorithm.The design of interface is divided into image area,file operation area and camera operation area.The implementation of the algorithm mainly includes usinginter-frame difference method for background modeling,using median filtering for image processed and using morphological approach for removing image noise.In addition,it includes binarization of image,selection of the threshold and so on.The development of the system uses the application framework structure of MFC and the class and library functions of OpenCV.And it adopts C++ object-oriented thinking to analysis and design,so that the system is not only have high stability and reliability,but also have good extendibility.The system realizes the detection of the moving objects in the warning area of the video file or the video information captured directly by the camera.When dangerous or critical behavior occurs in the marked alert area,the system will issue alert messages for the staff to respond in a timely manner.