Big data is promising in public security and other industries

In the subdivision of security, big data is applied earlier in public security and intelligent transportation exploration. The related solutions and technologies are also relatively mature. Related projects have also been launched in Guangxi and other places. Big data application system has been put into operation. Achieved the desired effect.

In the past year or two, big data began to be widely used in public security and other industries. In addition to the industry's own special requirements, big data has also driven the development of related industries. In the future, the industry application based on big data will become more in-depth, and more related companies will also be involved. The business model framework of big data in the public security field is gradually clear.

Big Data

In the subdivision of security, big data is applied earlier in public security and intelligent transportation exploration. The related solutions and technologies are also relatively mature. Related projects have also been launched in Guangxi and other places. Big data application system has been put into operation. Achieved the desired effect.

Project application prospects are promising

In related cases, in the big data system put into use by the Public Security Department of Guangxi, the entire project is based on the total data of the autonomous region, tens of millions of data generated every day in all aisles and about 3 billion a year. Strip data is distributed and quickly retrieved. On this basis, the follow-up can provide public security users with further solutions and value-added services, such as the already-launched solution for three major aspects of passing data, video image big data, and police intelligence big data. These programs provide multiple functions of query, and based on monitoring and control analysis and base station industry services, the purpose is to allow public security to quickly and scientifically detect cases.

In the field of intelligent transportation, it is mainly applied to the dredging of vehicles, such as statistics based on the traffic volumes of different roads and junctions (hourly, daily and monthly statistics, etc.). Based on these statistics, the traffic density and development direction of a certain road in real time can be analyzed. And trends and so on. The application of these projects has already landed in many major cities. For example, the real-time images of the rush hour sections that are broadcast on mobile TVs that are usually seen on public transportation are based on technical analysis of big data. From the application point of view, users really feel convenient and easy to use, so the market has great potential and future applications will be more extensive.

Big data application problems

Big data itself is valuable for data storage, retrieval, correlation, derivation, etc. These data are universal in nature. But in the field of security, what data is useful and what we need to care about and extract is what we are currently exploring. In other words, the current difficulty lies in how to combine technology hotspots and related businesses to extract more valuable data.

From the technical analysis, there are two technical difficulties:

The first difficulty is how to extract structured data from unstructured data. The so-called unstructured data refers to the extraction of features in the video. These may not be understood or processed by humans. Structured data is understandable and can be processed by people. For example, there are several activity targets in the video, and people are people. Still a car. If it is a person, what kind of clothes are worn on the body; if it is a car, how much is the license plate number, what brand model, color, speed, direction, and other data, these can be converted into structured data for human use. At present, many security data related to video data, and video data itself is not structured data, it can not be directly processed by the computer. Therefore, the topic that is facing the technicians in the future is how to convert video data into structured or semi-structured data that the computer can handle.

The second difficulty is finding the correlation and value between these data. The data is related and unrelated. We can only find it through tools. All of these stored characteristic data, including the massive amount of video data generated by the public security industry and Ping An City each day, can provide valuable clues for the investigation of many cases. The challenge that technology must overcome now is whether it is possible to correlate these data through relevant tool modules, link the original neglected data information through big data technology, and find or extract the correlation between these data. This is the case detection and decision-making. Provide scientific data basis.

The unidirectionality of data flow in public security

There are tens of millions of data obtained by the public security industry every day. How to ensure the security of these data information has become a common concern of the industry. In the traditional sense, after the data is generated, it is first necessary to ensure the security of the data itself. Currently, there are very mature technologies and solutions in the industry. In the face of huge amounts of data, if you do not understand the data, even if you put these data in front of you, it is difficult for you to extract useful data, but this can not be used as an excuse for the industry to ignore its importance. Because for security vendors, many valuable data need to provide protection, that is, adopt high-level protection measures for data application models. Once these data are tapped and associated by criminals, the entire region may have security loopholes. Will be used.

Now, public security data is generally run in local area networks and related protection measures are provided to provide security. If the data will be divided into different networks and different levels, the data will be unidirectionally flowed from the low security network to the high security network between different network security systems, and finally all the data will be collected in the public security core network. (This is the highest security level, protected by security boundaries, physical isolation). At the same time, the peripheral video network is mainly based on video data and supplemented with video-related services. These data can only be associated with other data after entering the public security network in order to discover some valuable data. For example, on the video network, you can obtain photos of criminal suspects. But who is this person and what is his information? Only after entering the public security network can he acquire it. Only by associating related information with other databases can you The association further digs out which internet cafe he has appeared in, and in which hotel he lived... The above information can be excavated, but this excavation can only be carried out in a high-security network. This information flow is one-way.

Future business model

From the traditional security business, it is mainly based on the public security client investment and construction system. The manufacturer provides products and integrated solutions, which are ultimately implemented by integrators, and finally delivered to customers for use and maintenance. At the same time, the future industry will become more professional in the acquisition, storage, analysis, and processing of data in big data. The users themselves may encounter various difficulties in processing and application. Therefore, there may be some small ones for such problems. Service companies emerged to provide end users with a variety of professional data services. For example, professional video extraction will be professional company cut-in, using professional algorithm tools to help you extract the data inside the video, or those professional communication vendors to mine and process the data, including the mode of providing some tools and services ( The future will be more inclined to the service model). However, due to the characteristics of the public security industry, these public services are still more difficult to do in the public security industry. However, in the future, some manufacturers can construct the entire application system and cooperate with public security clients or government agencies in the form of operating service charges.

For large-scale, particularly large-scale projects, such as projects involving a city, a province, and even a nationwide scope, generally speaking, they may find professional IT companies to do it. In particular, Internet companies (now involved), they are more With the participation of technology providers, security vendors focus on the business. In this way, the division of labor is relatively clear, because even if it is a technology leader in the industry, it is difficult or unnecessary to invest a lot of R&D in the research and development of big data, but should focus on the basic application of big data or business solutions. On the solution, the underlying infrastructure is then shared by IT vendors. Mutual benefit and mutual benefit with each other and sustainable development.