Industry News
    Home > News
  • The achieve and development of intelligent analysis of video surveillance

        In recent years, the popularity of the application of computer systems, network communication technology, image compression processing technology as well as the rapid development of transmission technology, security technology to prevent the industry to adopt the latest computer, communications and image processing technology, the use of computer networks to transmit digital images can be image monitoring and alarm system to provide efficient and viable solution. The development of video surveillance systems has experienced the first generation analog system, part of the second generation of digital system, the third generation of fully digital system (network cameras and video servers) three stages of evolution, the whole industry is brewing new video surveillance revolution - intelligent video surveillance.
        Intelligent video surveillance is a digitized, networked video surveillance, but different from the general network video surveillance, it is a more high-end video surveillance applications. Intelligent video surveillance system can automatically recognize different objects, found anomalies in the monitor screen and the fastest and best way to issue alerts and provide useful information, which can be more effective to help security personnel to deal with crisis, and to maximize reduce the phenomenon of false positives and false negatives.
        The concept of intelligent video surveillance
        Intelligent Video Surveillance (IVS Intelligent Video Surveillance) is based on computer vision technology to monitor the scene video content analysis to extract key information in the scene, and the formation of the corresponding event and alarm monitoring, is a new generation of monitoring based on video content analysis systems. Intelligent video surveillance technology with the powerful data processing capabilities of the computer, high-speed analysis of massive amounts of data in the video screen, filter out the users do not care about the information, only useful for monitoring critical information.
        Intelligent video surveillance system with functions
        1) the perimeter alert and intrusion detection: in complex weather conditions (such as rain, snow, fog, high winds, etc.) accurately detect and identify the movement of a single object or multiple objects and trajectory tracking, including the direction of movement , sports features;
        2) theft or motion detection; stolen objects in the surveillance scene and move, the algorithm will automatically detect this action, commonly used in the monitoring of the valuables and key equipment;
        3) left behind, abandoned object detection: When an object (such as boxes, parcels, vehicles, characters, etc.) in sensitive areas to stay too long, or more than a predefined length of time to generate an alarm. Typical scenarios include airports, train stations, subway stations.
        4) the number of traffic statistics: statistics through the entrance or designated area or object. For example, for the owners to calculate the number of customers that day to patronize their shops;
        5) congestion detection; identify the population as a whole movement characteristics, including speed, direction, and so on, in order to avoid the formation of congestion or to detect anomalies. Typical scenarios include a gathering place for the supermarkets, train stations and other personnel;
        6) PTZ tracking; detect moving objects, according to the motion of an object, automatically send PTZ control instructions, so that the camera can automatically track objects, the video surveillance camera continues to track objects beyond the range of the camera monitoring, automatic notification object Area ;
        7) fireworks detection to: - empty the characteristics of the real-time detection of pyrotechnics fireworks show when the fire occurred;
        8) human behavior analysis: target detection classification based on the use of a variety of behavioral characteristics of the human body description and analysis of its various acts extract what dangerous and potentially dangerous behavior, such as fighting, robbery suddenly fell places such acts;
        9) Face Recognition: automatically detect and identify a person´s facial features, and by comparison with the database file to identify or verify the identity of the character;
        10) Vehicle Identification: identify characteristics of shape, color, and license plate numbers of the vehicles, and feedback to monitor.
        Such applications can be used in the scene in the stolen vehicle tracking.
        Are several ways to achieve intelligent video
        IP video monitoring system based on computer network architecture, IP cameras or encoders for the front-end access equipment, network monitoring and management platform for the monitoring and management core equipment, PC client and decoder Browse backend video surveillance system. Based on the realization of intelligent video capabilities of the network video surveillance system is divided into the following ways:
        1, the server-side intelligence applications
        The so-called server-side smart, intelligent video analytics server dedicated These images are transferred to one or more video images, intelligent video analytics server from the network monitoring system or monitoring platform for intelligent rules setting produce intelligent alarm itself related business alarm or the Alarm business through the to reported network monitoring platform. Video intelligent business applications in this way, the front end of the monitoring system and the platform is only responsible for the intelligent video analysis required image resource scheduling to intelligent video analytics server, without any intelligent analysis of the work, the pressure on the front-end coding equipment is very small. And flexible scheduling image resources from the network monitoring system, the deployment of intelligent video business will be more flexible, ready to deploy at any point in the monitoring system, any network monitoring system through the system simple docking smart video technology.
        In large-scale monitoring systems, the server-side intelligent way to bring greater pressure on the back-end management resources, the need to alleviate by increasing the back-end equipment performance, such as improving server configuration to increase the number of servers. Usually experience, a high-performance quad-core CPU processing 16 CIF quality intelligent analysis applications.
        2, the video encoder DVS (or DVR) front smart
        The so-called front-end intelligence refers to video information directly through the front-end network video equipment intelligent analysis, intelligent video capabilities. In such a system, the front of the security surveillance camera is often achieved through the embedded video analysis services to analyze the captured video, the embedded system itself a powerful DSP chip processing capability during codec processing while using the processing resources of the DSP surplus intelligent video analysis. In this way will be accompanied by the strengthening of the front-end DSP processing power will increasingly become the mainstream way of video intelligence applications.
        Deployment of front-end intelligent front-end intelligent analysis of fixed point, specific and strong, stable performance, is the core business of video intelligent monitoring access methods in the future.
        3, front end, rear end, the combination of implementations
        This is a compromise, but also more reasonable and balanced intelligent video implementations. The one hand, make full use of the front-end network video equipment surplus CPU resources to achieve some simple intelligent video capabilities, on the other hand, the back-end video software to focus our resources to achieve a higher level or more application-oriented intelligent video capabilities, such as license plate recognition , behavior recognition, this system, intelligent video pressure more evenly distributed in the network video equipment and back-end management of the front-end server on each device to play his part, the system architecture is more reasonable. In the case of front-end DSP processing chip capacity is also unable to support a complete intelligent analysis business is a good way.
        Conclusion
        The intelligent, digital, network is the inevitable trend of the development of video surveillance, the emergence of intelligent video surveillance is the direct manifestation of this trend. Intelligent video surveillance equipment than ordinary network video surveillance equipment with more powerful image processing capabilities and intelligent factors, it can provide users with more advanced video analysis capabilities, it can greatly improve the ability of video surveillance systems, and video resources can play a greater role. Monitoring equipment, hardware suppliers, intelligent video software vendors, distributors and resellers and system integrators in order to promote the development of intelligent video industry to achieve a win-win situation effect of all these participants organized so that they can fully play to their strengths, to create a comprehensive solution to meet end-user needs.