The team develops an intelligent traffic linkage monitoring and analysis system based on intelligent video compression and retrieval technology, which can further improve the intelligent level of the traffic command system based on road surveillance video. Based on the latest artificial intelligence technologies such as pattern recognition and deep learning, the intelligent analysis of vehicles in surveillance video can achieve the purpose of highly intelligent traffic management. The system is based on the principle of surveillance video compression, according to the vehicle information in the surveillance video has temporal and spatial invariance, the vehicle in multiple cameras to achieve information association. Based on SSD neural network algorithm, the vehicle detection is realized, and on this basis, the vehicle color, model, speed, traffic flow and other information analysis are realized, so as to achieve intelligent monitoring of traffic information. The system is based on the adaptive background reconstruction algorithm based on exponential forgetting proposed by the team and a video summary generation method based on the key frame of foreground target, combined with the relevant principles of information and knowledge discovery in deep learning and big data, and realizes the vehicle information analysis based on the correlation between vehicle spatio-temporal information and different cameras in traffic monitoring video, it reduces the cost of human monitoring and greatly improves the intelligence of the traffic system based on road surveillance video.