What Is the Application Trend of Intelligent Video Surveillance in Cities
Smart city construction is not a simple slogan, but requires the actual implementation of system deployment one by one, and can meet the management and living needs of the government and the people according to the established requirements. In China, security technology is widely used in the construction of smart cities. For example, intelligent video surveillance technology has been extended to the business fields of various departments.
1〠Current situation and characteristics of intelligent video surveillance in smart city
As we all know, the intelligence of video surveillance has become an inevitable trend in the development of the industry. It can extract and screen abnormal behaviors in video in real time, and issue early warning in time, which has completely changed the passive state that traditional surveillance can only "monitor" but not "control". It will be a new revolution in the video surveillance industry. At this stage, more and more technology-based enterprises in China are vigorously developing core technologies such as digital signal processing and video analysis algorithms, which will greatly enhance the added value of the video monitoring industry and improve the development bottleneck of the difficulty in upgrading the value of the existing industry.
From a technical point of view, the current domestic intelligent analysis technology is mainly concentrated in two categories: one is to extract and detect the targets in the video picture by using methods such as picture segmentation and foreground extraction, and distinguish different events through different rules, so as to realize different judgments and generate corresponding alarm linkage, such as some of the earliest behavior analysis functions (cross border, regional intrusion, fight detection, personnel gathering, etc.) and early traffic event detection belong to the application of this kind of algorithm technology. The other is to use pattern recognition technology to model specific objects in the picture and train through a large number of samples, so as to detect and apply specific objects in the video picture.
From the perspective of application, there are four main categories of domestic intelligent analysis technologies: the first category is real-time alarm; the second category is data statistics; the third category is attribute recognition; and the fourth category is image processing.
The first category: real-time alarm. It mainly analyzes and judges the content of real-time video through analysis technology. When it is found that a certain state meets the requirements of alarm rules, the system can issue alarm linkage. For example, the most basic cross line alarm, intrusion alarm, fight alarm, crowd alarm, etc. of course, with the deepening of application, there are many industry specific applications in various industries The real-time alarm application of tax collection, such as campus fight alarm; traffic congestion alarm and pedestrian high-speed alarm in the transportation industry; climbing alarm and off-duty alarm in the judicial industry; trailing alarm and false advertising quotation in the financial industry, etc.
The second category: data statistics. It is mainly used to make statistics on the specific content of the video content in a specific scene to form relevant reports and data applications. For example, automatic statistics on the traffic flow on the road through video analysis, and statistics on the traffic flow at the entrance of the campus and shopping malls through video analysis.
The third category: attribute recognition. It is mainly used to automatically recognize the attributes of specific transactions in the video, so as to achieve the purpose of in-depth application and rapid retrieval of video content, such as face recognition, license plate recognition, vehicle logo recognition, color recognition, gender recognition, height recognition, age recognition, gesture recognition, etc. at present, face recognition and License plate recognition, vehicle logo recognition, etc.
The fourth category: image processing. It mainly analyzes, judges and optimizes the whole image to achieve better results, or the unclear content can be seen clearly through algorithm calculation and processing, such as the current video enhancement technology (denoising, defogging, sharpening, brightening, etc.) and video restoration technology (deblurring, distortion correction, etc.).
2〠Difficulties and bottlenecks of intelligent video surveillance in smart city application
The application of intelligent video surveillance technology provides convenience guarantee for safety supervision in various industries, and massive video data mining also helps urban management departments improve their decision-making efficiency and accuracy. For social public security governance, intelligent video surveillance technology not only brings unprecedented opportunities, but also brings many application bottlenecks and problems.
First of all, in the construction of video surveillance system in China in the past ten years, the original video surveillance system is independent, and the information island is serious. The public sector, private sector and other subjects involved in public security governance cooperate. Due to the lack of unified standards among various security data, the segmentation between existing organizations, departments and systems and the lag of information management concept often lead to The phenomenon of "data island" appears. Based on the application of intelligent video monitoring technology, the transformation of data application concept and the reconstruction of data management mode based on the basic principles of openness, transparency, sharing and cooperation will become the key to the application effect of video big data in smart city management.
Secondly, under the background of in-depth information construction of smart city, the existing video system has prominent problems such as lack of in-depth application mode and low intelligence of video data. Although various intelligent applications based on AI are in full swing in the application of video monitoring system, the current construction scale of AI based system is still relatively small, and its application bottleneck is the lack of video information standard Furthermore, there is a lack of a new police work mode of using video information to guide investigation and solve cases, and there is a lack of unified standards and norms in all links of video information and intelligence police application.
Finally, the essence of intelligent video surveillance technology is to digitize all kinds of things and integrate all kinds of data distributed in different fields, networks, systems and databases, so as to mine valuable information. In order to ensure the accuracy of information mining as much as possible, openness, circulation and transparency have become the inevitable requirements of video big data applications, especially in the field of public security In the era of big data of intelligent video surveillance, many smart city applications involve citizen property security and even national security, and the data value is very high. Therefore, the problem of information security in the era of big data has become a major problem in the construction of smart city.
3〠Application case of intelligent video surveillance in smart city
Jianggan District is the transportation hub of Hangzhou. It has the largest railway hub in Asia - Hangzhou east railway station, the highway passenger transport hub - Jiubao passenger transport center, and metro lines 1 and 4 with central radiation effect. Jianggan District is also the concentration area of religious places in Hangzhou, with Chongyi church, the largest Chinese Christian Church in the world, and the largest mosque in China. Jianggan District is also Special ethnic food gathering places, such as Pamir restaurant, Bayi master restaurant, etc. these places with large flow of people and unique background are all over the region. Under the background of the current threat of international terrorism penetration, they are more likely to attract the attention and attacks of terrorists and violent elements. The anti-terrorism and riot situation faced by Jianggan District, especially Qianjiang New City, is also more severe than that faced by other main urban areas.
In 2016, Hangzhou Jianggan branch launched the construction of intelligent big data management and control platform for video surveillance, focusing on police practice, using cutting-edge scientific and technological innovation applications such as cloud computing, big data, mobile Internet, Internet of things and artificial intelligence, and integrating data such as view resources, police resources and social resources, so as to fully tap the potential of data resources and build a platform for guiding police with data The public security organs have a new combat effectiveness, build a new normal of police application with big data characteristics, build a rigorous, ubiquitous, interconnected and intelligent multi-dimensional perception prevention and control circle, form a big data resource sharing and computing center with data washable, storable, mineable and predictable, and prepare a meta map that can display all kinds of resources, risk situation, patrol and defense areas, key places, etc Establish a joint service command center integrating the functions of intelligence research and judgment, synthetic operation and command and dispatching, form a new intelligent policing model of prediction and early warning, steady-state prevention and control, precision strike, synthetic operation, flat command, dynamic management and strong guarantee, and constantly improve the intelligent level of Jianggan public security work.
Since its launch in October 2017, the system has successively arrested more than 240 fugitives across the country, significantly improved the overall social security state of Jianggan District and greatly improved the satisfaction of the masses.
4〠Application trend of intelligent video surveillance in smart city
In the era of big data, with the continuous application of Internet of things, mobile Internet, computing and other technologies, the construction of smart city will vigorously promote the basic platform of big data, actively promote the construction of information resource data exchange and sharing system, and realize the integration of smart city and smart video big data. At the same time, with the continuous maturity of video big data related technologies, especially in processing The ability of structured data such as video and pictures has been continuously improved. Through various business models, behavior analysis, machine learning and other technologies, combined with massive data, the information in the video is intelligently analyzed and identified. The big data technology is used to establish a video based and open cloud service platform for the public, integrate more social resource information, and obtain support and ensure the smooth development of the smart city Benefit the diversified information of landing and operation. According to the user's business needs, provide video big data application services for social units such as finance, education and even ordinary consumers. Video big data based on intelligent video applications will become the key support for realizing "intelligence" in all fields of smart city.
epilogue
Smart city is a new generation of information technology to improve the information perception, information analysis and processing ability of urban managers in all aspects of the city, so as to improve the targeted intelligent whole society service mode. A large number of audio and video and other IOT sensing devices in the city need to be monitored in real time from key areas and key groups to carry out real-time early warning and prediction of suspicious situations In addition, we need to strike a balance between data mining and privacy protection to achieve real "prevention" and "security".
