Under the background of the rapid development of the Internet and big data, people's requirements for their living environment have gradually increased. Many consumers begin to pay attention to security products. On the one hand, they respond to the increasingly demanding needs of users, on the other hand, they also reduce labor expenditure costs, and the field of security monitoring system has gradually changed to intelligence. When more and more places popularize monitoring facilities, more and more places popularize HD monitoring, followed by massive data information and complex video retrieval.

Therefore, when intelligent analysis technology is applied in various industries, if it can distinguish the application environment and simplify the operation method, realize a single application, carry out specific development for each industry and embed special algorithms, or analyze only one or several simple events, such as personnel tracking at important entrances and exits, the system only needs to embed analysis and tracking algorithms, It will simplify the operation mode of intelligent analysis technology, and intelligent analysis technology will more fit the needs and characteristics of the industry and carry out more accurate analysis and operation.

On the one hand, the demand for industrial development of intelligent analysis technology comes from industry development and technical constraints, and on the other hand, it depends on the real feedback of practical application effect to a greater extent. At present, intelligent analysis technologies include behavior analysis, feature recognition, video diagnosis, classification and statistics, and the application focus of intelligent video analysis technology in different industries is also different. For example, the intelligent analysis system in the prison mainly includes cross-border detection, wandering event detection in the area, abnormal behavior identification, etc. it is mainly to prevent prisoners from escaping and group beating, and maintain perimeter safety. The application requirements for road traffic, expressway and other industries are mainly violation detection, traffic flow statistics, reverse driving, license plate recognition, traffic event detection and so on. For public places with high mobility such as airports and stations, the demand for abandoned object detection and wandering detection is more prominent in order to prevent the explosion of dangerous goods and the crime of suspicious personnel.

At present, intelligent analysis technology is mainly affected by the application environment, so many functions are not ideal in practical application. At present, the most common behavior analysis functions, such as cross line alarm, regional intrusion detection, personnel gathering, wandering and fighting, are seriously affected by the application environment and often fail to achieve the expected purpose in practical application. False positives or omissions are still on the high side. Moreover, the amount of information extracted from video is not enough for many deep-seated applications, and the judgment accuracy can not meet the requirements. However, we are still optimistic about the future prospect of intelligent monitoring. The monitoring system is still under construction every year. People alone can not handle the massive data generated by thousands of surveillance cameras. The intelligent analysis and in-depth application of video data is an inevitable trend. Now the intelligent application has reached the stage of industry segmentation. Through the actual project mode, some industries and scenarios suitable for intelligent analysis have been extracted for in-depth algorithm research. Many products have been able to achieve the purpose of practical use and solving customer problems. Moreover, with the continuous maturity of video analysis technology, more accurate analysis can be carried out for many complex scenes.

IVS has a bright future, but it still faces a long exploration journey. At present, two major problems need to be solved: 1. Solve redundant data, 2. Environmental impact, and how to break through intelligent analysis technology. The essential difference between intelligent video surveillance and ordinary video surveillance is that ordinary video surveillance only helps us "watch"; Intelligent video surveillance should not only "see", but also "see" and "understand", search targets, analyze video content and process video surveillance results.