"Smart Security" Makes Video Surveillance More Effective

In the 21st century, public safety is undoubtedly a hot topic of globalization. Its application has also been rampant - infiltrating all aspects of society, and it has begun to affect and change people's lifestyle. Security monitoring takes into consideration public safety, industry and commerce, transportation, and home and consumption. Applications for public safety have been written to cover the city's regional monitoring, government homeland security, and the public security law departments that are closely related to ensuring social stability; applications for industry and commerce have been rampant - from key facilities to conventional venues, making use of banks, games and Areas where people’s activities are frequent, such as retailing, are guaranteed to be safe; applications for transportation are constantly being innovative, and improvements in safety and security measures not only at stations, docks, airports, but also at roads, junctions, and fairways. .

Due to the importance of public safety, digital video technology and high-definition image technology are rapidly escalating under the impetus of the market and demand. Digital video monitoring standards have also grown geometrically. The situation is sometimes so interesting, digital video is becoming larger and larger, and the definition of video is becoming higher and higher. With the increase of cameras and the growth of storage devices, it cannot improve the effectiveness of video surveillance at the same time. The reason is very simple. The actual effect of all security systems depends on the integration of human defense, physical defense, and technical defense. The large-scale video surveillance scale and high-definition images cause the power series of image information to increase, which inevitably brings the operator's visual fatigue and difficulty in image retrieval playback, which affects the effectiveness of monitoring and tracing. It can be imagined that with the development of the Internet of Things in the future, the monitoring of the authenticity of objects will also extend to the scope of video surveillance, and the effectiveness of monitoring and storage of high-definition images in video surveillance systems will become more prominent.

The only way to solve the effectiveness of video surveillance is to make the camera have a higher "smart". From the visual experience of the human eye, it can be seen that the human eye sees a large-scale panoramic image including a 150-degree angle of view, and what is really clear is only the focus of the gaze. In the same way, the real need to pay attention to the monitoring screens of video surveillance cameras is not only the entire screen, but also a small part of them. If the camera can self-identify the focus of surveillance, and each time it transmits, displays, and stores this truly useful part of the picture, the effectiveness of video surveillance will be greatly improved, the transmission pressure of the image will be reduced, the storage capacity and storage devices will be Greatly reduced, the investment benefit of the project will also be greatly increased.

The biggest difference between intelligent security and traditional security lies in intelligence. China's security industry develops rapidly and is relatively popular. However, traditional security has a strong dependence on people and it is very labor intensive, while intelligent security can be intelligently determined by machines. May achieve what people want to do. Professor Yang uses intelligent video and non-intelligent video as an example. Intelligent video enables people to not have to stare at the screen and the machine automatically recognizes and can automatically report emergency events. Non-intelligent video surveillance requires people to judge.

Security of the Internet of Things also has intelligent judgments, such as perimeter prevention, natural conditions such as wind and rain, causing swinging of the fence, intelligent analysis can determine whether it is the influence of natural factors such as wind power, or there are events to exclude natural factors. The resulting false alarms occur, reducing the number of alarms caused by false alarms, reducing the labor intensity of security personnel, and greatly saving manpower. Smart Security is an event that handles security alarms according to the way people think.

China's security industry has encountered certain resistance in the process of moving toward intelligence. The most important resistance is the algorithm. Due to the different degree of intelligence in intelligent security, simple situations can be analyzed and judged, and it is not enough to meet complex situations. For example, the recognition of the flow of people and simple street flow recognition are also simple. However, in complex streets or complex scenes, intelligent video often cannot be identified during analysis, and the accuracy of recognition is low, and the accuracy of the algorithm may not guarantee the accuracy. For example, if the domain value of the perimeter system is high, it is easy to misreport, and if the domain value is low, it is easy to cause non-reporting. At present, the security level of the security industry is not enough, intelligence can not reach the level of human eye recognition, and the space for security and intelligent improvement is still large.

It goes without saying that in smart security projects, attention should also be paid to their relationship with existing technologies and products. Only when the new system is integrated with the original security system organically, it is possible to exert the benefits and benefits envisaged by the intelligent security. Expanding the surveillance scope of security to the information behavior of people forms "smart security", and there will be many new technologies and new problems that need to be studied and solved.