1、 The problem of format unification
The video involved in the case comes from a wide range of sources. In addition to the video resources from Ping’an City networking platform, there are also a large number of video resources from social monitoring points. Because the format and coding of videos from different manufacturers are not unified, the police often need to obtain the player corresponding to the manufacturer’s equipment model in order to access the research and judgment videos, and often the video cannot be played because the player cannot be found.
2、 Data collaboration
In the process of case detection, the relevant video data obtained by the police are often stored on their own computers for analysis, which can not be shared with others, which restricts the serial and parallel analysis of historical cases, and even leads to the omission of information when the staff work changes; In particular, there are many monitoring points involved in the case. When multiple people conduct image acquisition and collaborative cross judgment, the mode of decentralized storage of case video seriously restricts the efficiency of search, management and collaboration.
3、 Improving efficiency
The analysis and identification of crime surveillance video clues collected at the scene still stay at the level of visual observation and video playback, relying on manual discovery and intervention. When there is a lot of video data, the workload of watching video is very large and the efficiency is not high.
4、 Low application
The existing video image use level is low, only focus on real-time video and video information within the storage period, lack of intelligent, deep-seated associated application, so that a large number of video resources can not be used, which restricts the sustainable development of video investigation business.
Introduction to the scheme
Combined with the actual needs of the current public security for video investigation work, Dahua proposed the construction goal of the video investigation system of the public security organs, which is mainly based on the actual needs of the public security, video image network sharing platform, bayonet system, vehicle big data integration system and police business system Image processing technology and cloud computing technology, combined with the actual combat to explore a new video detection technology, through the video detection technology and other professional technical means organic combination, comprehensive application, the video detection work to achieve intelligent, practical, simple, efficient.
1、 Scheme architecture
The intelligent combat system is designed based on the advanced large platform architecture. Each public security business application and tool is not only an independent application (APP), but also shares the large platform system in the access of underlying devices and data acquisition. The system architecture design of independent APP based on large platform is designed to adapt to the continuous expansion of public security business functions, and provides a technical foundation for the system’s practical application, which is easier to develop two times. This kind of technology is easy to expand and develop the technology framework, and ensure the high degree of advanced and sustainable development capability of the large platform.
2、 Logical architecture
Access layer: realize the access and acquisition of all video resources and picture resources, such as public security self built security monitoring, bayonet monitoring, electronic police and wireless monitoring; Social monitoring resources include government agencies, banks, hospitals, schools and other self built monitoring resources, as well as private units, communities, shops and other monitoring resources.
Data layer: the collected data is stored in the same video image information database after structured processing, and is associated with other public security business data according to the actual situation of the case, providing a complete data view for each service to call.
Service layer: through various services, including unified transcoding, cloud storage, map service, intelligent analysis and engine analysis, the further processing and sorting of data can be realized, and the video, image and clue data can be effectively classified and managed;
Application layer: provide rich practical application methods and tools for case investigation, including case management, trajectory analysis, vehicle investigation, face investigation, video summary, snapshot retrieval, image enhancement, video restoration, wall map operation, video non editing processing.
3、 Scheme topology
Features of the scheme:
Dahua video investigation solution, based on the framework of “cloud storage, cloud computing” and the development concept of “instrumentalization”, focuses on the rapid investigation and easy use of video. It not only provides second level retrieval service of massive video image information database, but also brings the intelligent analysis and ease of use of the system to the extreme, fully meeting the demand of police wide application of public security video investigation business, It provides an important support for the realization of rapid detection and cracking down on crimes.
One minute to complete one hour high definition condensed summary processing
The innovative use of cloud technology, combined with intelligence, completely subverts the inefficient investigation pain points in the sea of people tactics, and achieves a new height in the application field of video investigation.
One person input, all staff share
Unified video image information database and fast retrieval service can realize second level sharing and search. The investigation information of a single police officer is collected and gathered to the view information database through the system. Other police officers with authority can quickly share information through keyword retrieval to maximize the value of view information.
One command, all in one
Based on the event, a multi police virtual case handling area should be established to break the barriers of time, space, resources and police services, quickly divide the work, share in time, cooperate with the whole police, and handle cases efficiently.
Complete tools and various tactics
It has more than 20 kinds of visual investigation tools and tactics in four categories of videos, images, vehicles and cases, which effectively assist the investigation work of investigators at all levels.
Close to the assessment, simple and easy to use, promote the popularization of the police
Statistical analysis of the construction progress, use frequency and application effect of map investigation from various dimensions such as unit, individual, system and equipment, and provide effective basis for assessment; At the same time, the system is simple and easy to use. Police of different ages and educational levels can quickly master and use it, which promotes the popularization of map detection technology in the whole police.
From actual combat to actual combat
The system has in-depth cooperation with local public security departments, and has a full understanding of the front-line investigation and case handling process; At the same time, Dahua, as an important compilation unit, deeply participates in the research on “key technology of video image information database” of the Ministry of public security, and is one of the important makers of industry standards and norms.
User value analysis
In the process of investigating and solving a case, through the comprehensive utilization of the huge image information resources of the video surveillance system, it can play its role in delimiting the direction of investigation, narrowing the scope of investigation, locking the suspected object, and even identifying the crime, which embodies the irreplaceable advantages of the traditional reconnaissance work.
1、 When a criminal suspect is found, the direction of investigation shall be determined
Using the visual and accurate characteristics of video surveillance, we can see the physical characteristics of the suspect directly, and successfully accomplish the transformation from “case to image, from image to person”, and provide an objective basis for the delineation of the correct investigation direction.
2、 Determine the law of crime and narrow the scope of investigation
Taking advantage of the “multi-point and wide-ranging” advantage of the video system, and using the recognizable clothing and other features on the image screen, the suspects are found. After the cases are concatenated, through the collection of multi period and multi-point video monitoring information, the time-space characteristics of the suspects are found by collision and comparison with the relevant information with long time-space distribution in the case prone areas, To narrow the scope of investigation.
3、 Identify the detailed features and comprehensively lock the objects
Due to the objective and continuous characteristics of video surveillance, the whole activity process of the suspect before and after the crime provides a wealth of information resources for the investigation and solving of the case. Through the analysis and judgment of a single frame image, the trace, fragment and characteristics of the suspect are determined, and the track information (such as telephone information) associated with it is obtained. A variety of investigation methods are comprehensively used, Direct targeting of suspects.
4、 Fixed video evidence, direct identification of crime
Video surveillance may reflect the whole process of committing a crime. The video recording directly the process of criminal activity directly fixed the suspect’s close-up in the time of crime, and identified the crime, which produced irrefutable proof.
Editor in charge: GT