Dynamic living face recognition parameters

Product introduction：

M1207 series face recognition integrated terminal uses the world's leading face detection and recognition algorithm (FDDB and LFW top three in the world), runs it in a high-performance embedded platform, with infrared, RGB dual cameras, dynamic dual photography to completely solve photo deception. The product can realize face detection, face tracking, and face recognition, while swiping the card, and can present the corresponding feedback on the screen. Face tracking and detection takes about 20ms, face feature extraction takes about 200ms, face comparison takes about 0.1ms, and it has good robustness to light and glasses. This product supports a variety of hardware interfaces, including HDMI HD output, serial port input/output, RJ45, USB interface; It has perfect software interface, including: equipment management, personnel/photo management, record viewing and other functions. At the same time, this product can interface with a variety of applications developed by partners, such as: ticketing, access control, visitor, person-ID integration, building management system, developers can also develop their own applications according to the interface of the product.

Product parameter

Dynamic living face recognition parameters

Interface: Provides secondary development interface and hardware device underlying protocol

Can access equipment: two-dimensional code reader, ID card reader, access control reader, etc
Application scenario:

Applicable to scenic ticketing, office areas, hotels, office buildings, schools, shopping malls, communities and other places that need to use face access control.
Product features:

l Support external large screen;

l Dynamic dual-camera live identification and anti-counterfeiting, completely solve the deception of all kinds of photos on various carriers;

l night infrared, RGB double complement light;

l Support wireless Internet access by inserting 4G network card;

l Support the storage of 100,000 face photos and 1 million recognition records;

l Support 1:1 face comparison, 1: N face recognition;

l 1: N recognition accuracy is high, when the face library is 10000 (it is recommended that the face library is not higher than 10000), the error rate of one in 10,000, 97% pass rate, 1:1 face comparison recognition accuracy of 99.7%;

l The recognition speed is fast, the face tracking and detection takes about 20ms, the face feature extraction takes about 200ms, and the face comparison takes about 0.1ms;

l Support stranger detection, stranger level can be configured;

l Support face recognition or stranger detection on-site photo preservation;