Intelligent monitoring can help deal with crisis

With the diversified development trend of video surveillance, the development momentum of intelligent video surveillance is also growing stronger. Intelligent video surveillance is the use of computer vision technology to analyze, understand and process video signals. Under the premise of operators providing corresponding rules, intelligent analysis Through automatic analysis of sequence images, the computer locates, identifies, and tracks targets in the monitoring scene, and analyzes and judges the behavior of the target based on this, that is, when an abnormal situation occurs, it can promptly issue an alarm or provide an alarm event information. Assist security personnel in handling the crisis.

The application of intelligent video surveillance technology in various fields has brought great help to our lives. For example, in the security of airports, railway stations and other places, through the video surveillance monitoring screen, intelligent analysis can detect dangerous people and target people in a timely manner, greatly improving the efficiency of crisis management; in home security, intelligent monitoring of the monitoring of the situation at home Can give warnings when there are deaf people invading, or when they are attacked by others, when they can't report to the police, video surveillance will automatically alarm when the screen analyzes to excessive behavior; in intelligent transportation, high-speed capture, license plate recognition and other urban intelligence The high IQ of video surveillance is reflected.

Although video intelligence analysis is powerful, its accuracy is still unsatisfactory. According to the current situation of technological development, video intelligence analysis is only a tool to improve the efficiency of monitoring personnel, and can not be a complete replacement for artificial computer advanced intelligence. According to the analysis of the reasons for its accuracy, there are the following main points:

Environmental factors such as the background chaos caused by sudden showers, changes in the target and background colors caused by light changes, and changes in the target and background resolution due to different conditions during the day and night are all factors that cause errors in the analysis. The uncontrollability of these factors also leads to the lack of stability of the intelligent analysis itself. The solution to this phenomenon remains to be further optimized by analysis algorithms and analysis techniques, and it still requires the continuous efforts of practitioners.

Hardware haunting hardware includes front-end cameras and back-end servers. For intelligent analysis based on pixel comparison, the sharpness of the image directly affects the result of the comparison, and the clearer the image points of the camera image are, the more the operation of the comparison makes the server more stressful. For example, license plate recognition in intelligent analysis, video resources obtained by cameras with different resolutions, and analysis accuracy rates vary greatly, and the comparison efficiency of back-end servers may also be significantly different. The analysis capability of a single server when using standard definition may be different. It is 20 lines and the accuracy rate is 90%. When using 2 million HD, the processing capacity may be 5, but the accuracy rate can reach 98%. This problem can only be solved when the front-end upgrading and "cloud computing" technology is effectively promoted.

Summary At present, although intelligent video surveillance is powerful, it cannot be compared with human beings. The basic reason is that intelligent video surveillance is based on human-controlled computers to determine the target's regular behavior. Many undefined rules can't be performed by the computer. Judgment, for example, in crowd gathering analysis, intelligent analysis can analyze the number of people in a certain area within a certain period of time has reached the expected value, but whether these people are occasionally passing or deliberately gathering can not be analyzed. This fully embodies the essence of computer intelligence analysis. It still cannot replace people's work fundamentally. This also just explains one principle: The most powerful thing is that people need to invent, create and operate applications.