In 2017, the “13th Five-Year” Specialized Planning for Technological Innovation in Public Security proposed “pushing forward the transformation from passive risk response to active risk prevention and control, from experience-driven to data-driven command and decision-making, from manpower-intensive to technology-intensive safety governance, and from case investigation and cracking based on clues tracking to targeted reaction.” As cloud computing, big data and AI technology grew mature with application in practical scenarios, more and more products conforming to requirements of this comprehensive trend were born out of need.
As the leader of global video monitoring management platform, NetPosa Technologies rolled out the more precise and effective video image analysis system based on its insight into industrial development trend and deep understanding of users’ need. The system was built on the basis of technologies such as multi-dimension sensing and collecting in the Internet of Things, intelligent structuralization analysis of videos and images, mixed cloud computing, big data cluster mining of videos and images, and complied with standards such as GA/T 1399 Application System of Video and Image Analysis for Public Security and GA/ T1400 Video & Image Information Application System for Public Security. It could analyze raw videos and images, and be combined with other sensing data (RFID, GPS) to provide integrated video and image information service which has real practicable value for the public safety field.
Provide big data service centered on people, vehicles and cases
The video image analysis system created by NetPosa Technologies provided intelligent video human face monitoring & analysis service centered on “people”. By offering intelligent, targeted and fast human face recognition and mature human face big data analysis and mining application, it realized real-time tracking of people under surveillance, quick comparison of people’s information, and backward searching of historical tracks, comprehensively solving problems of people management and surveillance such as searching people, early warning and tracking in the public safety field.
As well as people information, the system could also precisely and quickly recognize vehicles information such as their shape, color and license plate numbers. Vehicle big data service system centered on “vehicles” assisted in solving problems in intelligent city transport management application such as information collection, analysis & mining, and networking sharing of vehicle big data.
Meanwhile, video image analysis management service centered on “cases” could also be provided by processing and analyzing videos and images relevant to the cases to assist in solving problems such as clues tracking, analysis and troubleshooting, and cases connection so as to improve efficient use of resources and cases solving efficiency, and save manpower.
Transformative “searching videos” model raises cases solving efficiency
In the past, police officers had to search video recordings one by one to track clues, during which their cases solving efficiency was affected. NetPosa Technologies video image analysis system ended it by applying technologies such as big data, cloud computing, and AI to transform the traditional way of “watching videos” to today’s “searching videos”, which greatly improve the cases solving efficiency for police officers.
By taking full advantage of the cooperative computing ability of CPU and GPU, the video image analysis system applied the video image handling technology based on deep learning to automatically retrieve, recognize and have a structuralization analysis of real-time and previous video and image data resources. Meanwhile based on semantic analysis and video and image features comparison technologies, it could realize full-text retrieval of structuralization attributes of multi-dimensional video and image mixed big data of hundreds of billions of human faces, bodies, motor vehicles, and non-motor vehicles, as well as the searching-for-image-by-image function. Moreover, it could build video and image big data service system of tempo-spatial relationship, behavior relationship, and relevance relationship centered on people, vehicles and cases by combining multi-dimensional video and image resources and structuralized resources such as human faces, bodies, motor vehicles, non-motor vehicles, RFID and GPS.
At present, the NetPosa Technologies video image analysis system is applied in the project “video and image information integrated application platform of the Ministry of Public Security”, a top-level platform aimed for the construction of video and image application systems planned during the “13th Five-Year Plan”. NetPosa provides multi-level real-time service for the platform in the aspect of real-time video structuralization and intelligent analysis. In the future, NetPosa Technologies will continue to stay close with users, be creative, and forge ahead with the building of a stronger public security team with the help of technology and the establishment of a safe China.
