Design And Implementation Of Vehicle Video Surveillance System Based On Android Platform

With the continuous development of national economy and urbanization process, vehicles on the road are increasing continuously, the ensuing traffic problems are also increasingly prominent. Accordingly, the intelligent transportation for traffic management is growingly concerned by the whole society. The vehicle video surveillance is an important research topic in the field of the intelligent transportation, it can help users to visually monitor vehicle environment in real time.Traditional video monitoring generally adopts the way of fixing PC with proprietary network, which lacks flexibility and convenience. Recently, the solution of mobile video monitoring emerged with the popularization of mobile internet, however it has problems of poor video quality and low network resource utilization. Meanwhile, vehicle in the fast-moving needs better video transmission network.For solving the limitations of vehicle video monitoring system mentioned above, this paper introduces a research solution of vehicle video surveillance technology based on Android platform. First acquisition terminal encode collected data using H.264 and packet data on RTP protocol, then data is transmitted through UDP on the basis of hybrid network architecture based on P2P and C/S, and finally monitoring terminal unpack received data and use the FFmpeg to decode data in order to present video to the users.The research solution designed P2P and C/S mixed architecture to achieve effective method of going through NAT and mechanism of load balancing in order to improves network robustness and resource utilization; Using FEC、QoS algorithms to resist abnormal packets; Using multi-threaded mechanism to prevent thread blocking; Using the FFmpeg multimedia decoding framework and multi-cache strategy to improve the real-time monitoring performance.The experimental result indicates the system can achieve low delay, jitter and packet loss rate under different network conditions in the simulated environments, which meets the performance requirements of vehicle video monitoring system and is beneficial to extend to fields such as intelligent transportation.