Embedded Linux system, high performance AI processor, embedded deep learning algorithm
Support face detection and capture, face recognition, face optimization and face tracking
Binocular 10000 face database. The recognition accuracy is not less than 99.7%, and the recognition speed is less than 0.2 seconds
Professional ISP image processing, super intelligent, wide dynamic, fearless strong light, backlight and dark
2 megapixel, binocular camera live detection, standard 4mm lens, screen resolution 480 * 480
Night infrared / RGB dual complementary light, RJ45, RS232, switch output, Wigan, audio output
DC: 12V - + 30% power supply
Power supply: 12V1A
Product size: 15.6cm * 7.65cm * 15cm
Main board hardware: ARM7 + convolutional neural network accelerator NPU
Facial capacity: 10000
Offline record capacity: 20000
Background connection mode: LAN or public network
Display: 5-inch TFT color screen
Verification method: face recognition + live detection
Interface: switch output, Wigan output
Communication mode: TCP / IP
Light resistance: excellent
Display screen: resolution 480 * 80