Body temperature and face recognition technology is suitable to be installed for communities, office buildings, schools, hotels, scenic spots, transportation hubs, and other public service places.
These devices are designed to help prevent virus spread and maintain a healthy environment for your employees, associates, visitors, and the general public.
Facial recognition can be used to allow entry to pre-vetted individuals and/or store temperature readings for a user. Setting your own threshold for temperature readings, the device will give a successful message or warning with traffic light style LED lights.
The device is built using Rockchip RK3288 (optional RK3399/ Qualcomm MSM8953) high-performance CPU and is equipped with a commercial-grade binocular camera, live face recognition technology, and infrared thermal imaging module to support face-with-mask identification. The system supports automatic switching in 1: 1 and 1: N modes, and runs on the Android operating system. It has the features of face recognition, face-with-mask recognition, and human temperature detection, with automatic alarm for body temperature abnormality. It completes detection in seconds with high accuracy.
Accurate body temperature detection
- The kiosks support human body temperature detection and temperature display, with a marginal measurement error of ±0.2 °C (0.3 Meter) from close range, and ±0.5 °C from further away
- A distance of 0.5 meters is recommended for most accurate readings, with 1 meter being the longest distance at which temperature can be read
- It only takes a few seconds for detection and an alarm will sound automatic when temperature abnormality is detected – at a threshold set by you
Smart facial recognition
- Industrial-grade binocular wide dynamic camera, night infrared and LED dual photoflood lamp
- Face recognition pass speed is ~ 1 second
- Database capacity for 30,000 entries
- Supports recognition and comparison of faces with surgical masks on
- The 1: 1 comparison recognition rate is more than 99.7%, the 1: N comparison recognition rate is more than 96.7% at a 0.1% misrecognition rate, and the live detection accuracy rate is 98.3% at a 1% misrecognition rate