Artificial intelligence technology has enabled machines to "see" the people, cars, objects and behaviors in the video. The public security department can quickly find suspicious objects by comparing and retrieving data from the national fugitive library. So how does AI “seeing the eye†help the police solve the case?
In a surveillance video, basic information is displayed next to each pedestrian, such as male, middle-aged, wearing a cap, long-sleeved, and black pants. At the same time, there are basic information such as the model, license plate number, and body color next to the vehicle. Recently, at the "Ninth China International Police Equipment Expo" hosted by the Ministry of Public Security, many people stopped in front of this "video image analysis system" exhibited by Dongfang Net Power.
"The core of video analysis is to structure the video, extract the moving objects in the video stream through artificial intelligence algorithms, and give it certain attributes. At present, the structured display of face, human body, car and non-motor vehicle information has been realized. To achieve pre-warning, collaboration and post-event retrieval, and solve the most urgent dynamic face recognition needs of the public security department." Zhao Yongjun, president of Dongfang Netpower, said in an interview with reporters.
Manual case correlation analysis and trajectory tracking are easily interfered by factors such as subjective judgment, data analysis volume, and analysis dimension, which wastes labor costs and is inefficient. With the increase of video surveillance and the popularity of high-definition and networked in China, video is more than just a simple image. Through analysis and mining, more information data can be derived, which provides a solid way for artificial intelligence technology to enter the security industry. basis.
The reporter learned that China's security market scale reaches 100 billion yuan. The three major products of security equipment are video surveillance, access control and anti-theft alarm devices. Video surveillance is still the most important market. The artificial intelligence technology used in video surveillance is mainly face recognition technology, including static picture face recognition and dynamic video face recognition. The dynamic video stream is not structured and cannot be retrieved and utilized, and the video analysis developed to date has been manually The intelligent algorithm initially solved the structural problem.
Thanks to the real-time structured display of video, the Lianyungang Public Security Bureau of Jiangsu Province has greatly improved the efficiency of solving crimes. As of 2017, through the portrait tracking network, 35,860 key suspects were identified, and 76 fugitives were arrested. Through the vehicle tracking network, more than 1,100 cases were cracked, 685 criminals were arrested, and about 263 illegal vehicles were seized. It is estimated that in the middle of 2018, security will be more AI industrialized applications around the regional crowd monitoring and case analysis system.
The community will become the main battlefield of the city securityHowever, although China's video surveillance network has been gradually improved, but due to the installation time, technology, product form and landing scene, many cameras that have been controlled in public areas, truly meet the structural requirements of artificial intelligence, video analysis is still not much . Zhao Yongjun believes that the large-scale application of products suitable for the scene at the front end is an important development direction.
Some people say that the community is the longest place for human activities, and people living in the house produce a lot of living data. Insiders pointed out that the community will become one of the main battlefields of urban security. In Zhao Yongjun's view, community security must create three lines of defense. Each line of defense has different concerns, but it is inseparable from artificial intelligence technology. The first is the community door, focusing on the data collection of personnel vehicles in and out. The second is to enter the entire road from the community to the door of the unit, through the video capture and other data collection to achieve the trajectory of the target. The third way is to realize the connection between the unit door and the access control system, that is to say, the future access control system has the face recognition function. "Through three lines of defense, you can master a variety of data such as the number of people entering and exiting, identification of personnel, and the presence or absence of trailing."
At present, many security companies have developed intelligent security community systems, which are inseparable from the dynamic sensing data of community monitoring, access control, face recognition and other systems, thus facilitating the collection and management of various basic data in the jurisdiction. . In addition, some companies provide programs in security early warning, such as Vanke's dangerous area entry warning, anti-lost management system and early warning of child abnormal behavior management.
AI security is the general trend of security industry upgrades. However, the industry also pointed out that how to combine the collected security scene data samples with actual combat, how to make the identified data move from shallow information to deep information, how to improve the security field monitoring data. Openness and degree of sharing are challenges in the application of artificial intelligence technology in the security industry.
ALL Kinds of 4G Cat-4 Vehicle GPS Tracker, there must be one which should meet your expectation.
ES810-C4 the one supports WIFI and Hotspot
ES811-C4 the waterproof version of ES810-C4
CAT-4 vehicle GPS Trackers,CAT-4 car GPS Trackers,Waterproof CAT-4 vehicle GPS Tracker,IP65 CAT-4 vehicle GPS Tracker
eSky wireless Inc , https://www.eskygpsiot.com