-GB/T28 18 1 Technical Requirements for Information Transmission, Exchange and Control of Security Video Surveillance Networking System
-GB/T 26718-2011Technical Requirements for Urban Rail Transit Security System
-Participate in the compilation and development of the National Video Surveillance Standards Working Group (SVAC)
Regional/industry standards
-As the only safety manufacturer, participate in the research of six major topics of "safety engineering"
-National Grid Standards; DB33, Zhejiang Province; Testing standards of Ministry of Railways; Global eye standard; Anhui Provincial Public Security Department interconnection standard; Sichuan electric power interconnection standard; Internet standards for electronic inspection; "Key Technologies and Equipment of Large-scale Metropolitan Monitoring Streaming Media" such as Wide Vision Standard won the second prize of 20 10 National Science and Technology Progress Award.
Project name: Key technologies and equipment of large-scale subway monitoring streaming media.
Mainly completed by: Chen, Ji Xiangyang, Wang, Yu Furong, Ding Guiguang, Duan Huilong, Tian Xiang, Jiang Rongxin, Ma,.
The main completion units are Zhejiang University, Tsinghua University, Ningbo University, Hangzhou Huasan Communication Technology Co., Ltd. (Storage and Multimedia Division is independent of Shiyu Technology in the following year) and Nanchang Hangkong University.
At the beginning of this century, with the increasing demand of international anti-terrorism and domestic social security, China started the major project of "Safe City 311"to support the construction of large-scale metropolitan monitoring system. Because it involves national security infrastructure, monitoring technology and equipment must be based in China. Large-scale subway monitoring refers to the real-time monitoring of important places and activities in the city by using streaming media equipment. Different from the general streaming media transmission and processing technology, the data transmission scale of subway monitoring system is large, for example, the streaming media data flow of100000 monitoring devices is 10tb per minute, which requires high real-time performance and is transmitted in the form of high concurrent streaming media. Realizing efficient encoding and real-time analysis and processing of concurrent streams is the key to solve the technical problems of massive streaming media.
In this environment, Shi Yu Science and Technology cooperated with many well-known universities in China to set up a large-scale subway monitoring technology research project. With the support of 973,863, Natural Science Foundation of China and national key scientific and technological projects, the project took the lead in tackling key technologies of large-scale subway monitoring system, and made breakthroughs in key technologies such as adaptive scalable coding, strong fault tolerance of concurrent streaming media, multi-source and multi-scale video retrieval and virtual network storage (which is a popular cloud storage technology in recent years), forming a number of technologies and equipment with international standards. The project applied for 47 invention patents, of which 32 were authorized and the software copyright was granted 13.
The main innovations and technical indicators are as follows:
1. In order to solve the problem of efficient coding of high concurrent surveillance video under the conditions of heterogeneous networks and various terminal access devices, a region adaptive coding method is invented, which integrates the subjective quality evaluation feedback mechanism of visual perception model, breaks through the limitation of single source rate distortion model, and constructs an efficient scalable coding method of PTSQ, which saves more than 30% of the code rate compared with unicast method, and is accepted by the reference software of international standard (ISO/IEC MPEG).
2. Aiming at the problem that it is difficult to realize high reliable transmission of multi-channel video concurrent streams, the prediction error transmission law of video streams in the case of packet loss is revealed, and the temporal and spatial correlation of motion vector fields is described. A multi-channel video error control and joint error recovery method based on distributed source coding is proposed, which can improve the peak signal-to-noise ratio of reconstructed video quality by 0.81db-1.10% under the packet loss rate.
3. Aiming at the analysis and retrieval problems caused by multi-source multi-scale and time correlation of surveillance video, a real-time feature extraction method of compressed domain multi-channel video is proposed, a correlation feedback learning mechanism and semantic labeling strategy of interactive retrieval are constructed, and a multi-label nearest neighbor propagation retrieval method is invented to realize real-time analysis and rapid retrieval of massive surveillance video.
4. Aiming at the limited concurrent processing of massive video streams, a large-scale subway monitoring system architecture with virtual network storage is proposed, which breaks through the bottleneck of real-time scheduling of massive streaming media in the original hierarchical media server system architecture, develops a variety of extensible real-time monitoring devices, and constructs a large-scale subway monitoring system.
As one of the important participants in the formulation and implementation of "Safe City 311"by the Ministry of Public Security, Shi Yu Science and Technology has made important contributions to the popularization of large-scale subway monitoring system. The developed system has been successfully applied to more than 200 safety projects in Beijing, Shenzhen, Chongqing and other places, playing an irreplaceable role in the construction of safety projects and making outstanding contributions to the development of IP video surveillance technology in China.