The Junpeng Jiemu facial recognition product is based on the Ruixin Micro RK3288 high-performance quad core application processor platform. The RK3288 main chip integrates quad core Cortex-A17 and Mali-T764 high-performance GPU, with a maximum clock speed of 1.6GHz. It has super computing performance, 2D/3D graphics processing capability, and full HD video encoding and decoding capability, perfectly supporting 4Kx2K@60fps Ultra clear decoding and 4Kx2K HDMI ultra clear output.

Features:

1: The accuracy of facial recognition exceeds that of human eyes, and recognition can be completed within 1 second (the facial feature library can be optimized with a multi-threaded binary tree algorithm, and even 100000 faces can be easily achieved in milliseconds);

2: 2 million to 3 million high-definition wide dynamic cameras, with live detection function enabled, can recognize distances up to 6 meters;

3: Equipped with monocular liveness detection, binocular liveness detection function, and embedded software and hardware integrated design;

4: 7-inch TFT high-definition full-color screen, compatible with horizontal and vertical display interfaces;

5: Prevent hacker attacks on 3D printing, electronic screens, masks, and headgear;

6: The system comes with built-in daemon processes and performance reset to ensure continuous and uninterrupted operation (with system running status logs for easy management and maintenance in the future);
7: The system has strong maintainability, centralized authorization in the cloud, convenient for group management, and the equipment can support both online and offline scenarios;

8: Supports a whitelist of 30000 faces, with a recognition comparison time of less than 1 second, and supports exporting recognition logs;

9: Support batch facial input, deletion, and information modification methods across all platforms: front-end devices, PC management software, cloud facial recognition, mobile APP facial modification, etc.

10: Support remote modification of system settings across all platforms, such as remote modification of system language, output signals, recognition methods, recognition distances, etc;

11: Support multiple combination verification methods, such as face verification, face+swipe card verification, face, swipe card, QR code verification, etc;

12: Support adding validity period and number of passes for facial data;

13: Unique facial management interface, easy to operate, low performance consumption, and strong scalability;

14: Widely practical, with multiple output interfaces: switch output, Wiegand signal output;