Design Of Multi-channel Video Recording System And Software Development Based On Linux

In order to train the algorithm model,a special image recorder is needed to collect a large number of pictures and video data in deep learning technology for visual classification and detection.Traditional consumer video recording equipment usually only supports single-channel shooting,while professional video monitoring equipment has some problems including poor portability and inconvenient operation.To solve these problems,by research and comparison of technical scheme,this thesis designs and actualizes an embedded multi-channel video recording system software based on AM5728 multi core processor.The software of this equipment is integrated with multiple-shooting,good mobility and convenient operation.This thesis main contents are as follows:(1)The overall design of multi-channel video recording system based on Linux,not only expounds the research status of home and abroad,introduces the related technologies used in this thesis briefly,but also summarizes the software and hardware environment as well as the structure of this thesis.(2)The multi-process information interaction system is used to design and implement the communication framework between the client and the server based on the TCP network protocol.The data transmission and communication functions are continuously optimized for WIFI and 4G different network environments.The inter-process communication method of socket and shared memory are adopted to realize data transmission,parameter query and data save in the system.(3)In order to improve the portability of the multi-channel recording system,the Gstreamer streaming multimedia framework is used to record multi-channel video,and the related plug-ins are used to construct a multi-channel video recording system,which is encoded by H.264 for local storage.The perceptual algorithm is used to eliminate the redundant components in the image sequence.The serial communication protocol between the modules of the system is designed to realize the data transmission.(4)In order to share data and automatically expand in real time,the data is stored in the Alibaba Cloud server by means of network transmission,the video data and log files in the system are uploaded by means of slice uploading and append uploading,which realizes the real-time sharing and management of data.(5)Summarize the research content of this thesis and look forward to the part that the system can improve.