Video cloud service technology advantages and impact on security monitoring
Guangdong Rixin Precision Technology Co., Ltd. , https://www.nissonmold.com
Cloud Video is a video streaming service based on the concept of cloud computing technology, covering all processes from acquisition to playback, enabling customers to build professional video systems in a cost-effective and efficient way. In the cloud video service process, the video collected by the content provider is first encoded into a specific format, after which the video is uploaded to the cloud server, transcoded in the cloud to adapt to different playback servers, and finally, the video is accelerated via a content distribution network (CDN). Distribute, after decoding, play in the terminal.
Video cloud services eliminate the need for physical storage and maintenance. Therefore, it can save investment in physical security technology and reduce investment in total assets. As the core video coding technology in video cloud services continues to advance, data compression reduces data volume, saves storage space and costs, and reduces the use of network bandwidth. Accelerated integration with technology makes CDN efficient, low-cost, and intelligent. The evolution of the direction; video cloud combined with artificial intelligence, can provide users with more diverse value-added services.
First, video cloud services have the following obvious advantages over traditional platforms:
(1) Video image information is more comprehensive, information analysis is faster and more accurate
The cloud service platform is equipped with a cloud storage system, which has powerful centralized storage capacity of massive information, supports more than 1000 storage nodes, and has a capacity of one hundred petabytes. It can be transparently expanded online by means of horizontal expansion. After expansion, the original data is automatically redistributed to All node clocks balance the storage load and the calculated load.
Since these video image information are centrally stored and unified into the cloud storage system, the retrieval, fusion and joint analysis of the image information will be more comprehensive and faster, and the comprehensive processing capability of the information will be stronger. The effect is directly reflected in the actual operation of the public security actual combat department. It is faster for image information retrieval, video summary analysis, comprehensive information analysis and judgment, and more information will be applied and comprehensively analyzed. accurate.
(2) Video surveillance integrated application is more efficient and flexible
In the cloud service platform, various types of resources are centrally managed, and real-time resource dynamic monitoring and scheduling can realize on-demand scheduling of service resources and have powerful resource management scheduling capabilities. Therefore, the cloud computing system has stronger image processing service parallel processing capability, more flexible service capacity expansion capability, and faster service deployment capability than the system under the traditional architecture.
Through the centralized management and flexible analysis of virtual resources, the cloud platform can realize the isomorphization and measurability of computing resources, support the next generation technology mode of dynamic resource pool, virtualization and high availability, so that it can provide as small as one computer. , the computing power of up to a thousand computers. Users only need to know what kind of computing power they need, what time they need to have these computer skills, and the rest are managed and scheduled by the cloud platform.
Based on the virtual resource pool, the cloud platform supports dynamic resource scaling and network redundancy of the basic resources, which means adding, deleting, and modifying any resource node of the cloud computing environment, or any resource node is abnormally down. , will not lead to the interruption of various types of business in the cloud environment.
In addition, the dynamic flow of resources within the cloud service means that the resource scheduling mechanism is implemented under the cloud platform, and the resources can be transferred to the required places. For example, when the overall system service is increased, idle resources can be started to be included in the system to improve the carrying capacity of the entire cloud service. In the case that the entire system has a low business load, the services can be centralized, and other idle resources can be transferred to the energy-saving mode, thereby achieving the green and low-carbon application effects of other resources while improving the utilization of some resources. Under this feature of on-demand resource allocation, public security video integrated service resources can grow with the increase of workload, and will not encounter performance bottlenecks and affect business execution.
(3) The entire business system is smarter and more powerful
Because cloud services have the functions of parallel distributed computing processing and automatic expansion of computing resources, they have powerful information analysis capabilities, which can not only support multi-type information fusion analysis, but also have higher real-time performance for information processing. Relying on these capabilities, in the security video surveillance application, it is possible to gradually realize the transformation of the post-prediction prediction and analysis.
In addition, due to the stronger information analysis capability, the system can process video information more efficiently and accurately. In intelligent video analysis, the video image information can be summarized and analyzed more accurately, which can be more quickly and accurately. The characteristics and activity trajectories of the development of related targets can also achieve a lower intelligent analysis false positive rate and improve the accuracy of early warning. Based on the big data analysis and processing basis provided by the computer platform, various kinds of clue information and other structured and unstructured data after video analysis and processing can also correlate and analyze faster and more accurately, and support various complicated techniques of the public security department. Innovative application of warfare.
Second, the benefits of video cloud for security monitoring
In recent years, the video surveillance system has been continuously constructed and developed rapidly, and it has played an increasingly active role in combating crime, public security, social management, and serving people's livelihood. Judging from the deployment of the current national video surveillance system, the video cloud construction model has become a star.
Information service provider IHSMarkit released the white paper "Video Surveillance: Technology and Cloud is Changing the Traditional Market", pointing out that video analytics will increasingly be implemented through distributed architectures - smart surveillance cameras Edge computing, converged back-end servers and cloud-based central video analytics, enabling video analytics to be applied to a wider range of real-world scenarios to support real-world applications.
Video cloud computing is based on the concept of cloud computing technology, using video as a cloud computing solution that presents the processing results to the "terminal" from the "cloud". The application runs on the cloud server, encodes the running display output and sound output, and transmits it to the terminal in real time through the network. The terminal displays the output after real-time decoding. The terminal can operate at the same time, and the operation control information is transmitted to the cloud application running platform for application control in real time through the network, and the terminal is "streamlined" to provide only network capability, video decoding capability and human-computer interaction capability.
With the integration of security and Internet, Internet of Things, cloud computing and other technologies, the emergence of video surveillance HD cameras, the security industry is becoming digital, integrated, networked, intelligent, and civilian. The application requirements of cameras in various industries have evolved from simple security protection to high-definition, intelligent and intelligent remote visualization management.
The white paper points out that cloud architecture will be widely used in safe cities. Cloud architecture has tremendous advantages, including: video data sharing between regions and across organizations. Serve different organizations with a unified platform to reduce the construction of duplicate ICT facilities. Resources can be managed more efficiently and flexibly assigned according to the requirements of the task. The cloud architecture is designed to provide an open application support platform that provides system integrators and end users with more choices. The white paper also highlights the importance of the ecological environment. Traditional video surveillance vendors use vertical E2E mode, which greatly limits the user's choice of algorithms and applications.
3. What benefits will video cloud services bring to traditional security monitoring?
(1) Improving the efficiency of resource use
The resource pooling of computing and storage can greatly improve the efficiency of resource utilization, realize time-sharing, cross-domain sharing, and on-demand calling of resources, avoid redundant construction, and realize rapid and flexible resource expansion and accelerate service deployment.
Let's take two examples of typical scenarios:
Scenario 1: In the morning and evening traffic peaks of the city, the video cloud can allocate more computing resources to the secondary identification of the bayonet, while in the off-peak period, more resources can be allocated to key areas (such as the railway station). , personnel control of key areas.
Scenario 2: Under the traditional situation, the new service of the video surveillance system has a long online schedule and cannot meet the rapid growth requirements of the video surveillance service. After the video surveillance system is on the cloud, it is very convenient to expand existing services and launch new services. The cycle is reduced from half a year to weekly.
(2) Providing the possibility of data fusion
Facilitate the integration of video surveillance data between cities, districts and counties, form a "data lake", realize video surveillance image information sharing, and make video tracking operations across cities, districts and counties possible.
Under the new situation, the National Development and Reform Commission, the China National Comprehensive Development Administration, and the Ministry of Public Security and other nine ministries and commissions jointly issued the No. 996 document in 2015, which proposed "global coverage, network-wide sharing, full-time availability, and full-process control" for video surveillance systems. The requirements for the construction of a three-dimensional prevention and control system.
In order to meet the requirements of "full network sharing", the traditional video surveillance system is built using a chimney mode, and video data sharing between different manufacturers' systems is very difficult to implement. The cloud platform is a mode for solving data sharing. The video cloud is used to carry large image data of video surveillance, which has the advantages of smooth expansion, easy sharing, easy processing, and easy control.
(3) Solving the problem of computing power of massive video image information big data and AI processing
Video surveillance image information Big data and AI processing and analysis require ultra-high performance computing capabilities, and the cloud can solve the problem of insufficient computing power.
Massive video image information needs to go far beyond the storage, analysis, and search capabilities of common data processing technologies. Video big data can analyze indirect human-to-human, object-to-object, and human-to-material relationships, which poses a very high challenge to the computing power of video surveillance systems. The video cloud solution provides computing power on demand for video surveillance systems. For example, a province's big data platform, powerful computing capabilities to support rapid response to business, can achieve billions of data analysis in 1 minute. There was a landslide and a vehicle was buried. At that time, through the big data platform, the information of 75 toll stations, 4042 kilometers of high-speed in the province, and 133,000 vehicles were combined, and information such as the vehicle model, occupants and quantity of the buried vehicles was accurately analyzed in combination with information such as time and space trajectory, which provided strong support for the rescue work. The traditional server cluster solution can not meet the above requirements.
(4) An open cloud model to build a prosperous ecosystem
The cloud platform is an ecology, which can be docked with a variety of camera cameras from different manufacturers to carry the industry-wide algorithms and applications. Through flexible combination, we always maintain the leading edge of actual combat capability.
The use of a general-purpose server to build the infrastructure is the basic capability of the video cloud. On top of this, a cloud platform based on OpenStack open architecture and a big data platform based on Hadoop architecture are built as a unified security cloud computing storage environment, realizing software and hardware solutions. Decoupling, front-end and back-end decoupling, application and platform decoupling. Supporting the deployment of various algorithms and applications, allowing each vendor to do what they are good at; providing reliable, open, easy-to-maintain, easy-to-manage unified access services and video transcoding services, compatible with various mainstream vendors. The front-end equipment meets the business needs of the public security organs for daily three-dimensional prevention and control.
The video cloud is the platform and ecological basis. Open and sharing are essential genes. In a certain province, the innovative component development model is used by the police to develop more than 100 common components, which are convenient for each police to directly call according to business needs, improve efficiency and achieve excellent practical results.
(5) Cloud intelligence is more powerful and richer than front-end intelligence
The core of video cloud construction is the sharing of all the cameras. All camera and video data resources should be on demand. For example, the main job of urban road bayonet is vehicle identification, but it is also necessary for important security or emergency arrest operations. Face recognition function, all algorithms cannot be built in the front-end camera, and multiple types of cameras need to be deployed to meet the demand. It is better to implement multi-service algorithms from the cloud.
The five driving forces enable more and more video surveillance systems to adopt cloud mode for planning, construction and deployment, realizing resource sharing, data fusion, and computing power; while open ecology and business intelligence enable video surveillance systems Better serve the public security, make the city safer and make the society more stable.