“As of September this year, the number of video surveillance cameras installed for public service in the country has reached 30 million units, and the detection rate under the support of video image technology has been increasing year by year.†Tan Xiaozhun, Party Secretary of the Ministry of Public Security Science and Technology Information Bureau, stated that in AI's empowerment Next, the intelligent analysis and sharing application technology of video and image information can be said to be even more powerful in the field of security.
With the development of technologies such as cloud computing, big data, Internet of things, mobile Internet, and artificial intelligence, the detection of modern public security systems has long been more than just relying on the visits of criminal investigators with super-strong brains. Advanced artificial intelligence security products such as face recognition technology, VR+ video technology, etc.
“As of September this year, the number of video surveillance cameras installed in the country for public services has reached 30 million units, and the detection rate under the support of video image technology has been increasing year by year.†Tan Xiaozhun, Party Secretary of the Ministry of Public Security’s Bureau of Science and Technology Information, said that the empowerment of AI Next, the intelligent analysis and sharing application technology of video and image information can be said to be even more powerful in the field of security.
Video Surveillance Construction and High-tech Co-frequency Resonance "The characteristics of video images are intuitive, accurate, timely, and rich in information content. They have the characteristics of anti-terrorism, stability, command and outreach, public security prevention and control, investigation and detection, social management, law enforcement supervision, and service for people's livelihood. A very wide range of applications. At the same time, the video image technology support rate has been increasing year by year.†Tan Xiaozhun cited an example. Zhejiang used the public security portrait matching system only for 3 months, compared with more than a thousand criminal suspects, video surveillance. The efficiency of squeezing criminal space has also become increasingly apparent. Since the construction of video surveillance in Jiangsu, only cases involving theft, plagiarism, and theft of non-motor vehicles in homes have fallen by 19.9%, 18.2% and 16.9% respectively in 2016.
“This year’s significant development of video and image work has been an important experience in continuously strengthening technological innovation and actively promoting the application of new technologies such as cloud computing, big data, internet of things, and artificial intelligence to ensure the highest level of video surveillance application and application. The advanced technology is developing at the same time, with the same frequency resonance.†Tan Xiaozhun said.
“At present, our country has already been at the forefront of the world in terms of technology, standards, construction scale, quantity, and networking status of video surveillance systems. The corresponding technical standards system is also basically formed.†Intelligent analysis and sharing application technology of video image information Chen Chaowu, director of the National Engineering Laboratory, told reporters that the General Office of the Ministry of Public Security combined with the Secretariat Office of the Central Comprehensive Management Office, the General Office of the National Development and Reform Commission, and the National Standards Committee Office issued the "Public Security Video Image Information Sharing Application Standard System (2017 Edition)" The technical standards, test specifications and management specifications have been clarified. These three categories have a total of 24 national standards and provide basic technical support for the inter-regional, cross-industry, and cross-departmental development of the public security video surveillance security construction network. In addition, the Ministry of Public Security has also released six public safety industry standards such as video surveillance platforms, databases, and interface protocols so that public security agencies can deeply apply video image information.
Massive video data needs to be fully utilized
The wide range of valuable content that the video involves involves excavating the value, providing the most intuitive, accurate, and abundant clues for social security, providing great convenience for the prevention of cases and rapid detection. Through intelligent excavation, advance prevention and precise strike of cases can be better achieved, effectively supporting the integration of public security in fighting prevention and control. However, massive video technology also brings certain challenges.
"On the one hand, the number of video surveillance installations in the country has exceeded 30 million, resulting in massive amounts of video data. The value is still far from being fully tapped and utilized. On the other hand, emerging technologies such as cloud computing, big data and AI are evolving with each passing day. How to use these technologies to achieve deep application of video remains to be further explored, said Yu Bing, deputy director of the First Institute of the Ministry of Public Security and general manager of Beijing Zhongdun Safety Technology Development Company.
It is reported that the National Public Security Image Information Ministry-level networking platform developed and produced by the First Research Institute of the Ministry of Public Security has built a public information network and video special network cross-network parallel ministry, provincial, city, and county-level grass-roots bureau teams, and multi-level multi-domain interconnected national video. Image information network sharing architecture, with millions of video equipment supervision, has now connected more than 30 provincial video platforms across the country, and has more than 80,000 video surveillance devices.
"Because surveillance cameras generally have blind spots, according to fragmented images, it is impossible to effectively grasp or analyze the overall situation of the entire region, and it is difficult to use a large amount of data, which causes great confusion for command and decision making." Professor, Beijing University of Aeronautics and Astronautics Week Zhong Zhong said that as of 2015, 65% of China's big data are surveillance videos, and surveillance cameras are still increasing at a rate of 20% each year. The videos taken by each camera are relatively independent. In this case, the demand for video fusion emerged.
A few years ago, Zhou Zhong and other researchers broke through the unilateral modeling technique, built the location of the scene fragments first, and then reconstructed a complete scene model. "This technology is compatible with the currently operating monitoring system. If only one server is installed at the back end and a VR terminal is installed at the front end, the back-end content from different storage devices can be integrated to provide an integrated front-end display. "Zhou Zhong said.
Deep learning makes security more intelligent. "The growth of data is very impressive. There are many sources of data, many types of data, and strong correlation between data. As far as big data itself is concerned, it is not big. It is the value generated and the Cost, these are more important than the data itself." Chen Chaowu said that even valuable data tends to exceed the cognitive limit of human beings. When mining related data, it is necessary to use artificial intelligence to push data with the need; Is the need to use information display equipment, a new type of presentation.
Huang Kaiqi, a researcher at the Institute of Automation at the Chinese Academy of Sciences, believes that intelligent video surveillance technology is to make computers like human brains, let the eyes of camera heads, intelligently analyze the image sequences acquired from cameras, and understand the contents of monitored scenes. , to achieve automatic warning and alarm of abnormal behavior.
For the security industry that has mastered many video image resources, the combination of deep learning and security has a relatively high degree of fit. In image analysis, such as familiar face recognition, text recognition, and large-scale image classification, deep learning has greatly improved the accuracy of complex task classification, and has greatly improved the accuracy of image recognition, speech recognition, and semantic understanding. In terms of human face, face detection, face key point location, identity card comparison, clustering, face attributes, and live detection can be achieved. In terms of intelligent monitoring, it can be used as a structural study of video of people, motor vehicles, and non-motor vehicles.
"In-depth learning has enabled security systems to gradually change from being mainly human-based views to machines for viewing and analysis, and has achieved a leap from simple viewing to true prevention and control." Qiu Zheng, vice president of Beijing Zhongxing Microelectronics Co., Ltd., stated that the deep learning algorithm is By learning a large amount of data, the algorithm automatically obtains no longer relying on manual intervention and the existing identification model. The neural network represented by deep learning is very suitable for applications in the field of security and video surveillance, mainly reflected in: first, high accuracy; second, it can be used in the background server and front-end camera application mode, and no longer requires the transplantation of this software. Development; Third, adaptability, can be customized for specific goals of learning; Fourth, flexible deployment can be achieved to meet the needs of a multi-purpose machine.
“We are currently facing the needs of public security centers and various policing operations. We are developing video big data storage, processing analysis and mining technologies, and promoting the development of public security video integrated information intelligence analysis and judgment platforms and public security image detection integrated application platforms. Data technology, to explore the depth of intelligent analysis and sharing of video image information and big data applications and research." Yu Bing said.
Gear reduction motor is a component of a gear reducer and a motor. This type of component is also commonly referred to as a gearbox motor or a geared motor. It is usually assembled and assembled by a professional reducer manufacturer.
Gear reduction motors are widely used in automatic mechanical equipment, especially in packaging machinery, printing machinery, corrugated machinery, color box machinery, conveying machinery, food machinery, three-dimensional parking equipment, automatic storage, three-dimensional warehouse, chemical, textile, dyeing and finishing equipment, etc. Micro gear reduction motors are also widely used in the fields of electronic locks, optical equipment, precision instruments and financial equipment.
Gear reduction motor,Reduction motor,Speed reducer motor,12 volt gear reduction motor,Gear reduction box electric motor,12v gear reduction motor
Shenzhen Maintex Intelligent Control Co., Ltd. , https://www.maintexmotor.com