Product Brief
Facial recognition adopts the internationally advanced neural network algorithm (CNN), which is a product formed after tens of millions of algorithm training. It integrates functions such as image acquisition, face detection, face tracking, and face comparison, with high recognition rate and fast recognition speed. Offline dynamic portrait recognition machine can directly recognize pedestrians without the need for special cooperation or computer control. Pedestrians can also be recognized while in motion, greatly improving the usability of portrait recognition. The product can be applied to any channel gate (wing gate, swing gate, three roller gate, etc.), access control and advertising door commonly used in the market, and can directly output the opening signal and Wiegand signal; It can be widely used in home entrances, community access control, school entrances and exits, barrier channels, office buildings, subways, airports, high-speed trains and other scenarios.
Product Description
At present, the most commonly used method for entrances and exits in residential areas and factory areas is card verification, which often leads to the problem of not carrying a card. Facial recognition technology is applied to entrance and exit gate systems, which is a more secure identity verification method. Our facial recognition cameras can be applied to access control systems, with advantages such as high recognition rate, fast recognition speed, and adaptability to high and low temperature harsh environments. After successful facial comparison, the camera can directly output switch signals to devices such as turnstiles, access control, and advertising doors to complete the door opening action.
Technical Specifications:
1. Abnormal body temperature voice alarm;
2. Paired with a 7-inch LCD display screen;
3. Temperature detection error: ± 0.4 ℃/indoor
4. Temperature detection distance: 0.3-0.5 meters;
5. Working environment temperature: -40 ℃ to+80 ℃
6. Equipped with a 6mm large aperture lens and intelligent fill light;
7. Aluminum alloy material, embedded Linux operating system
8. Front end capture, comparison, recognition, output opening signal, no need to configure server;
9. DC12V3A power supply, tablet weight 2.2kg, cylinder weight 3.5kg.
10. Low light CMOS, 0.01Lux @ (F1.2, AGC ON);
11. Recognition distance: 0.5-3 meters; Recognition speed ≤ 250ms, recognition rate ≥ 97%;
12. Support panoramic and close-up image output (successfully compared, not compared, failed compared, synthesized image);
13. Support temporary list control: set the validity period of the face, which can be precise to the time point, and it will automatically become invalid if it exceeds the time point;
14. Facial recognition can be directly captured on the camera without the need for a facial recognition device, and can also be done using mobile phone photos or inch photos; Batch upload all faces
15. Support 20000 portrait libraries; Identification type: 1: N; supports video overlay; Support the Wiegand protocol; Support black and white list and alarm output;
·16. Equipped with electronic shutter function: personnel do not need to stand, pause or wait, do not need to deliberately cooperate with recognition, even when running, they can recognize, truly achieving dynamic recognition function;
17. It can be used offline and online. One computer can connect to 16 facial cameras, and multiple computers can also be connected to the internet. The software includes functions such as adding, modifying, deleting, querying, attendance, and visitors;
18. 2 million pixels, using embedded systems, HiSilicon industrial grade chips, Sony 1/2.8 "wide dynamic low light sensor device, image resolution 1920 (H) * 1080 (V), frame rate up to 25 frames per second;
19. High performance embedded processor, integrating image acquisition, face detection, face tracking, face comparison, and live judgment, can recognize in indoor and outdoor, day and night, cloudy, rainy, dusk, direct sunlight, and low light environments;
