Design And Implementation Of Multi-channel Surveillance Video Information Management System

With the popularization of video surveillance and the improvement of technology,more and more value of surveillance video remains to be discovered.The development of deep learning technology makes the processing of images more efficient.How to realize the storage and playback of many surveillance videos has become a key issue.This thesis sets up a thesis to study the multi-channel surveillance video scheduling and storage system based on the analysis of the driver's intelligent video behavior.It uses cluster technology,network technology,multimedia and other technologies to form a final solution,which realizes the efficient storage and multi-channel surveillance video.Dynamic scheduling.This thesis first analyzes the background and research purpose of the project,and then clarifies the research content of this thesis in conjunction with specific investigations on domestic and foreign surveillance programs,video storage programs,and video playback programs.The related technologies required for the solution were sorted and introduced in detail.Next is a research and analysis of the actual needs of the project's multi-channel surveillance video scheduling and storage system,with the help of use case diagrams,sequence diagrams,activity diagrams and other tools to gradually improve the specific functional requirements and performance requirements.Next,the general design of the system was carried out according to the determined requirements,and displayed in a hierarchical architecture diagram,etc.,as the overall framework for the next specific implementation of the system.Furthermore,the functions of the database,client and server of the multi-channel surveillance video scheduling and storage system are implemented.Finally,all the function points of the system are tested and some test cases and test results are listed.The innovations of this thesis include: 1.Video display directly realizes the playback of video from HDFS,reducing the cost increase caused by multi-server clusters;2.Load balancing distribution of video playback requests to solve the bottleneck problem caused by single server.