YITU Dragonfly Eye™ Intelligent Security System is built upon machine vision technology to operate an intelligent security system designed for safe cities, intelligent transportation and smart finance.
Like the myriad of small eyes that render dragonflies one of the most visually perceptive animals in the animal kingdom, YITU sets up a Generic Portrait Platform and a Generic Vehicle Platform based on world’s leading machine vision algorithms and platform solutions to provide accurate visual perception in the world. These City Eyes offer accurate recognition and detection to help build reliable a platforms for safe cities, intelligent transportation and smart finance.
YITU Dragonfly Eye Generic Portrait Platform, built upon our proprietary face recognition technology, can be used for static facial detection as well as dynamic facial detection and comparison of video stream. It has now been widely applied to public security systems in many provinces in China.
YITU's Dynamic Portrait System is designed for face recognition under dynamic video streams. It incorporates such functions as face retrieval and recognition, real-time surveillance, and passers-by tracking and enquiry. It has been widely applied in such crowded areas as residential districts, commercial zones, railway stations, subway stations, bus hubs, airports, customs, and entrances and exits of shopping malls. Under the surveillance and alarm module, anyone on blacklists entering monitored areas will trigger an alarm. Under the passers-by tracking and enquiry module, suspects can be traced to identify their movement pattern and their accomplices and provide important visual and location information for law enforcement.
YITU Dragonfly Eye™ Intelligent Security System is built upon machine vision technology to operate an intelligent security system tailored to cities and industries by means of video monitoring facilities designed for safe cities, intelligent transportation and smart finance.
Like the myriad of small eyes that render dragonflies one of the most visually perceptive creatures in the animal kingdom, YITU sets up a Generic Portrait Platform and a Generic Vehicle Platform based on world’s leading machine vision algorithms and platform solutions to provide the most accurate visual perception in the world. These City Eyes offer accurate recognition and detection to help build reliable a platforms for safe cities, intelligent transportation and smart finance.
Our Static Portrait System supports large-scale portrait comparison. Available for 1:1, 1:N and n:N, it has been widely used in public security and financial areas. It can be used to build a database for urban population and selected groups to quickly and accurately identify the person under check through portrait comparison.
City monitoring videos
Metro monitoring videos
Public monitoring videos
Security system pictures
Sampling population database
Static population database
Dynamic portrait engine
Duplicate-checking of population database
Surveillance and alarming
Offline video analysis
Personnel composition analysis
Behavioral pattern mining
Hiding place analysis
YITU’s Dragonfly Eye Generic Vehicle Platform is the cutting-edge intelligent car-recognition system in China. It was included in the Annual Recommended List of Science and Technology Achievements by the Ministry of Public Security in 2014 and won the Science and Technology Progress Award from the Ministry of Public Security in 2015. The Platform boasts proprietary capabilities of car brand identification, fake license plates analysis and “car search by image” based on visual features. YITU outperforms its peers in the generic vehicle platform technology in terms of technical parameters. Many core technologies are exclusively developed and owned by YITU Technology.
1. Car identification in images and video stream
2. Car brand/sub-brand identification in front and rear
3. Efficient and accurate analysis of fake license plates
4. Car search by image from massive images and videos
Security system pictures
Motor Vehicle Administration database
Video security system
Two-step recognition database
Video security system database
Static recognition engine
Dynamic recognition engine
Car search by image engine
Fake license plate
Trucks on highway viaducts
Car search by image
Occupying bus lane
Surveillance and alarming
Offline video analysis
Hunting and seizure of escaping suspects with disguised license plate
First-time entry analysis
Hiding in daytime and coming out at night
Hiding place analysis of suspects
YITU’s Dragonfly Eye Generic Portrait Platform technology has been widely applied in identity authentication, fugitive hunting with composite images, investigation and detection, and citizenship services in the public security system. Through our platform, the police can timely authenticate the identity of anyone under investigation, confirm unknown identity, re-locate suspects in disguise that cannot be recognized by traditional investigation means, trace suspects, conduct real-time monitoring, sound alarms on key persons, help missing elderly people back home, and provide Internet-based identity authentication services for the general public.
Every day, customs and ports in China are faced with intense pressures of the two-way flow of people and goods. Manual tracking methods are unable to meet their security needs. YITU’s Dynamic Face Recognition System is able to efficiently and accurately detect illegal activities and safeguard the country’s borders.
Each year, illicit goods crossing the country’s border illegally gross tens of billions of RMB, resulting in substantial loss of tariffs and duties taxes. YITU’s Dynamic Face Recognition System can accurately locate smugglers and suspects in huge crowds, inform law enforcement, deter the criminals, and help maintain the country’s tariff integrity.
Finance institutions previously conducted real name verification by means of manual recognition, text message verification and bank card linking. These traditional methods are unreliable and costly in addition to generating poor customer experience. Real name verification based on face recognition, which has already been adopted by many leading financial firms, is proven to have extremely high accuracy (as low as two identical in 100 million people), good customer experience (fast verification and simplified process), and low cost (compared to traditional methods). Bringing an integrated solution of real name verification based on facial comparison and vivo detection technology, YITU offers comprehensive services for financial firms covering all channels of bank counters, mobile terminals and ATMs. Our key clients include China Merchants Bank, Shanghai Pudong Development Bank, JD Finance, 360 Finance and many other financial firms.
In brick-and-mortar businesses, expenditures on target customer identification and targeted marketing constitute a major operating cost. As such traditional means of targed marketing are losing ground, targeting based on AI technologies has become a new frontier for business growth.
Cooperating with Wanda Plaza, YITU offers tailored services to VIP customers that are automatically identified by our technology solution. Personalized content is pushed to customers in real time based on their business needs in terms of gender, age and other important attributes. These attributes are generated by machine vision through identification and generalization of facial features to guide target customers and conduct targeted marketing. Moreover, pushed content will better match the needs of target groups through observation and learning the content of interest of different groups.
Through long-term and in-depth cooperation with the Suzhou Public Security Bureau, YITU Dragonfly Eye Generic Portrait Platform technology has been favorably recognized through all-round application in their public security system. Utilizing YITU’s Generic Portrait Platform, Suzhou police has launched a proactive investigation approach that identifies suspects in disguise using large scale matches. Three fugitives were arrested on the first day when the system was put in operation. Since June 2015, Suzhou police has successfully cracked over 500 cases and arrested 169 suspects.
YITU Face Recognition System was officially put into operation in the Xiamen Police Command Center on November, 2014. Since then, data of more than 2.5 million people has been imported into Xiamen’s permanent resident database, key person database, criminal database and other key databases. Using the face recognition system, Xiamen police has detected over 30 suspects, arrested 18, and cracked down on 60 burglary, robbery and pick-pocketing cases. Xiamen Water and Land Public Transport Sub-bureau, an affiliate to Xiamen Police Command Centre, has used image clues in their investigation. As a result, pick-pocketing cases on buses fell by 30 percent and the number of pick-pocketing cases dropped from 27 to 6 per day during the National Day Holiday in October.
Capitalizing on YITU’s advanced technology of facial recognition and the capacity for large data analysis, Wuhan Image Investigation Division has built a generic portrait platform that assists in investigation, decision-making, detection and seizure, and development of multi-dimensional solutions that provides early warning when any signs are spotted under suspicious circumstances. In the first three months after the system was put in operation, 132 cases have been solved, setting a stellar example of integrating face recognition technologies into police detection practices.
The city of Ningbo's face-recognition system was officially put into operation on October 9, 2016. During the G20 Summit, Ningbo Rail Transit Administration Sub-bureau, an entity under Ningbo Municipal Public Security Bureau, arrested 9 suspects at large. By deploying YITUs dynamic face recognition platform technology at key traffic junctions in the city, Ningbo police has made full use of many functions of the YITU system including face capture and recognition, real-time surveillance and passers-by inquiry that help trigger alarms once a suspect on the blacklist enters a monitored area, trace the suspects movement, and identify their accomplices to provide assistance to the police for further investigation.
Violation of traffic law is a major cause of fatal traffic accidents. In early 2016, Fuzhou traffic police adopted a system supported by YITU’s technologies to record traffic light violation by pedestrians and non-motor vehicles, aiming to curb traffic violations. YITU’s dynamic face recognition technology allows the system to take automatic and real-time photographs of pedestrians and non-motor vehicles that run red lights at the city’s major junctions in a quick and accurate manner. The system then automatically reads and records the offenders’ identity and other basic information as material evidence for warning or fines.
Sixteen years ago, Zhang Liwei (former Abbot of Longxing Temple), together with his accomplices, killed three people and then escaped. Zhang Liwei changed his identity and under a new name Sun Hongtao, obtained legitimate residential status at Mazhai Police Station, Yingzhou District, Fuyang Municipal Public Security Bureau. He moved to Langya District, Chuzhou in 2009 and was promoted to the Abbot of Longxing Temple in 2014.
In August 2016, using YITU’s face recognition system, the Criminal Investigation Brigade of Longzihu Sub-Bureau under Bengbu Municipal Public Security Bureau spotted facial similarity between a suspect from northeast China Zhang Liwei and the Abbot of Longxing Temple of Fengyang County Sun Hongtao. After further investigation, Zhang Liwei was detained and interrogated on site where he confessed his crimes.
In August 2016, the Disqualified Drivers Traffic Violations Analysis System, with the core algorithm engine developed by YITU Tech, was officially put into operation in Shanghai. It is designed to use data screening and high-accuracy analysis to reduce the number of unlicensed drivers on the street.
The System will first screen the drivers’ central database to obtain the information of any cars registered under the name of the “disqualified” driver from the motor vehicle database. Data on these drivers passing all traffic junctions during the disqualification period will be retrieved by their motor vehicle license plate. Then the System will analyze and compare the photograph on the driver’s file and that of the driver in the car, using the portrait and face recognition technology to determine whether a disqualified driver is driving.
Once confirmed by the System, information of the "disqualified" driver and their car will be sent to all police checkpoints in Shanghai. When such vehicles try to enter or leave the city, the System will give early warning for police interception.
A suspects' physical features may be hidden or altered and their physical features changed over time, making it increasingly difficult for investigators to capture fugitives. In late 2015, with the help of YITU’s portrait system using n:N screening, former bank branch director of Laishui County He Dongsheng, who embezzled 1.5 million yuan and has been at large for 15 years was finally arrested and prosecuted after he was located at a residential block in Jinan Shandong using portrait screening and analysis.