Research And Application Of Key Technologies Of Campus Security Based On Video Analysis Technology

As the bottom line of school development,safety is one of the core of campus construction.With the gradual rise of smart campus,smart campus is no longer a concept,but gradually transformed into a wide range of applications in reality.Among them,security is the core of smart campus.At present,video monitoring has become the standard configuration of school security construction.However,the video monitoring system mainly stays in the real-time viewing and playback of objective scenes,and the application of video security recognition based on artificial intelligence and other technologies is still very low.How to use artificial intelligence,video recognition and other technologies to establish an intelligent campus video monitoring and recognition system is an urgent problem for intelligent campus.Aiming at the shortcomings of the current campus video monitoring system.Taking campus security as the starting point,based on video analysis technology,this paper studies the key technologies of video vehicle target recognition,smoke automatic detection,pedestrian detection and so on in campus security scene,and designs and realizes campus security system based on video analysis technology.The main work and achievements of this paper are as follows:(1)Aiming at the shortcomings of traditional smoke detection in accuracy and real-time,a smoke detection method based on optical flow improvement and Yolov3 is proposed.Firstly,the dynamic foreground region is selected by the improved optical flow algorithm,and then the primary screening results are input into the Yolov3 model for secondary identification and screening,so as to achieve the purpose of smoke detection.The experimental results show that the model can effectively reduce the interference of external factors and complete the smoke detection task.(2)Aiming at the deficiency of single feature extraction in pedestrian recognition statistics,a pedestrian detection and statistics algorithm based on multi feature fusion is designed and implemented.Feature fusion and feature selection optimization are carried out for multiple features.SVM classifier is trained by positive and negative samples,and pedestrian recognition is carried out for the input image.Finally,the mean shift tracking algorithm combined with Kalman filter is used to make statisticsand analysis of pedestrians in the set area.(3)A campus security system based on video analysis technology is implemented.The system combines the above research results,realizes the integration of smoke automatic recognition,pedestrian detection and statistics,face recognition,vehicle recognition and other security modules,and establishes a campus security system based on video analysis technology.The experimental results show that the campus security system based on video analysis technology proposed in this paper can realize the transformation from civil defense to technical defense.It has better practical value and popularization value.