In the meeting of twelfth Public Security Expo on November 1st, when talking about  the monitoring of video technology for public security work, Ministry of Public Security, Dr. Merlin, from the 3rd Institute said that at present, with the constant expansion of monitoring networks and the mature applications of surveillance video, the existing public safety video surveillance system can not fully meet the needs of police officers, front-line police put forward higher requirements for the performance and the technical means of video surveillance network, which is mainly manifested in three areas:
      First, there is a gap between image quality and the needs of handling cases. For this proble, in addiction to have the quality needs of some acquisition camera terminal improved, one important reason is to save storage space and transmission costs, often using video image codec with large compression ration, it will cause serious loss of detail, and in handling the cases, many useful information get lost in data compression.
      Second, monitoring network lack of application in intelligent technology. The image content analysis of video surveillance network still need to rely on artificial real-time viewing and manual reading. With the expansion of monitoring networks, massive growth of video data, image viewing and reading occupied large police force, it becomes performance bottlenecks in making full use of video surveillance network. It is more and more difficult to get useful information from mass of surveillance video. In consideration of actual application, information should be retrieved and inquired, and the current applications of video surveillance monitors are mostly scenes of reproduction and playback, it is hard to meet the modern needs of public security information technology applications.
      Third, video data mainly transmitted over special monitoring network, and can not truly be shared by public security private network. Monitoring network are equipped with very large number of cameras, neither real-time viewing or afterwards video file check, public security network can only access a small part of video resources, and limited by network bandwidth, The application can not be shared in large-scale. Similarly, video surveillance data of other sectors of society also can not be accessed and shared by security network users. This results in the problem that a large amount of video data isolated on an information islands, can not quickly and effectively reach the hands of police handling the cases.
      Dr. Merlin said, the above problems, the Third Institute of Public Security Ministry proposed a structured description of video surveillance technology solutions. Video structured description is to study how to retrieve the video (unstructured data) and other forwards in-depth application technology. The key technologies of surveillance video structured description include five terms: video abstraction, image feature extraction techniques, video intelligent analysis technology, ASIC technology and embedded devices, structured description standards of surveillance video.