The working principle in the application scenario of the combination of human and witness: firstly, the face image or video stream is collected by the front-end camera, and the system can automatically detect and track the face in the image or dynamic video, through a series of related technologies, face location and quality detection are carried out on the collected images.
* Finally, cross-compare the pictures obtained on the spot with the extracted identity document information to ensure the consistency and legality between the holder and the certificate. On 2016, the Group of Twenty (G20)
Leaders 11th times summit will be in, China Zhejiang Hangzhou held in more a security Mount reference face recognition technology-assisted or instead work personnel to complete the stay personnel of identity verification work. It is the successful performance of the combination of witnesses.
On April 21, 2016, the newly installed witness system in five security card points in Korla city, Xinjiang was officially put into use after testing.
Third, there are many kinds of identification documents, just like Wentong express card can identify ID card, driving license, driver's license, Hong Kong and Macao Pass, ID card can clearly identify the ID information, it is convenient for the hotel to keep customer information.
Certificate identification supports 180-degree and 90-degree automatic rotation. Certificate identification supports complex backgrounds (If you take the certificate in your hand)
The edge-cutting certificate identification supports the automatic tilt correction function, which improves the recognition rate. The post-processing of the certificate identification result adds rules to the identification result, and improves the recognition rate. The shooting of the certificate identification specification is helpful to improve the recognition rate and illumination, pay attention to the influence of light when shooting, try to avoid reflection and shadow;
Angle, do not tilt the shooting angle too much, so as not to cause serious image variation; Background, leave less background (That is, the certificate is full of pictures)Or a simple background can improve the recognition rate;
Certificate identification support platform Android2. 3 more than, iOS6.
Above 0, windows 32/64 (
Windows Service 08, 12), Linux32/64 (
Ubuntu, CentOS, RedHat)
Certificate identification supports secondary development and provides Android development JAR package and iOS platform.
A static library development kit, which provides various interface modes such as WebService, Rest Service, Http, etc. Document recognition mode the first video preview mode uses scanning document pictures, and then scans the document information in the document Pictures (
Such as name, gender, nationality, date of birth, address, ID number, etc)
And generate document text information and avatar pictures.
Fourth, face recognition technology is based on human face features, for input face images or video streams.
Firstly, it is judged whether there is a face. If there is a face, the position and size of each face and the position information of each major facial organ are further given.
According to this information, the identity features contained in each person's face are further extracted and compared with known faces, thus identifying the identity of each person's face.
Five, authentication and authentication, face recognition, is an inevitable choice for security under modern technology, and the identity verification system based on face recognition technology can work for days without affecting efficiency.
It avoids the forgery and fraudulent use of information, increases the security of information, makes the combination of witnesses more reliable, the combination of witnesses more flexible, and the combination of witnesses more rapid.
The application architecture of the witness verification system is mainly divided into two types: 1. The witness comparison and the gate trigger part of the personnel to brush the second-generation ID card, and extract the FACE image information in the second-generation ID card, and with the person's face information collected by the high-definition camera, the recognition degree is compared. If the recognition result is correct, the gate will be opened to allow pedestrians to pass.
2 The part of face collection and database building is mainly to model and store the identity information and face information of people through photo information or face camera collection information.
Through the camera on the face recognition terminal, the face information to be passed through the pedestrian gate is collected and compared with the personnel information in the library. If the recognition result is correct, the gate will be turned on and allowed to pass.
If there is no person's face information in the library or the special certificate is not enough, if the comparison is unsuccessful, the gate can be opened in other ways.
The main function of the witness verification gate is to capture face photos in real time on the spot, compare them with ID card photos and information, and then enter the public security control network for identity verification, implementing one person, one card and one checkpoint, strictly control the passage of illegal personnel.