The Design And Implementation Of Embedded Intelligent Monitoring System Based On Moving Object Detection

With the gradual deepening of intelligent video surveillance and the increasing processing capacity of embedded chips,the demand for research and development of embedded intelligent monitoring devices has been unprecedented.The intelligentization of embedded monitoring equipment can effectively reduce the cost of video surveillance systems and improve the monitoring efficiency of the system.Based on the research of object detection and abnormal behavior recognition in video surveillance,this thesis designs and implements a set of embedded intelligent monitoring system based on B/S software architecture.In this thesis,based on the moving object detection algorithm in the video,the optical flow characteristics and image energy that can characterize the behavior are calculated,and then the intelligent detection of illegal intrusions,rapid passage,and violent events in the monitoring scene is implemented in the embedded platform.The main contents of this thesis are as follows:1.Completing the selection of development tools and building the development platform.Firstly,based on the analysis of the functional requirements of the system,we completed the design of the overall architecture of the system,then selected the important tools in the system development process,and finally completed the construction of the system hardware and software development platform.2.Research of Moving Object Detection Algorithm in Surveillance Video.Detection of moving objects in video is the key step of anomaly behavior recognition.The detection effect of the moving object directly affects the accuracy of subsequent abnormal behavior recognition.Among many moving object detection algorithms,ViBe algorithm has a small amount of computation while obtaining good detection results,so it is particularly suitable for the moving target detection algorithm on the embedded platform of this thesis.However,the original ViBe algorithm has “ghosting” phenomenon and is sensitive to light mutations.This thesis solves this problem effectively by combining the ViBe algorithm with the inter-frame difference method.Finally,the detection accuracy of the algorithm is improved at the expense of less time complexity.3.Detailed design and implementation of the system software.The system software is developed by B/S architecture,which can divided into two core parts,one part is responsible for video image processing,the other part is responsible for controlling system and user interaction,The above two parts also contain multiple functional modules.This thesis has carried on the detailed design and the realization to these modules.4.System testing.Firstly,seting up the laboratory's internal test environment,and testing the system's function and performance according to the standard test method.The test results show that the embedded intelligent monitoring system designed in this thesis can meet the requirements of system development.