Nine million. In-depth learning of smart cameras
The VT-E309 .9 million pixel in-depth learning smart camera was independently developed by BEIJING VION TECHNOLOGY,INC. It is the first high-definition smart camera in the industry to use Global Shutter COMS sensors, support in-depth learning, 4096 × 2160@25fps video output, and internal integration of multiple sensors. The product adopts dual chip processing architecture, and the processing performance is greatly improved. The product uses a 1 "large target Global Shutter COMS sensor, which has better imaging effects, lower power consumption, and better low-illumination effects, and can effectively remove the Smear phenomenon. The depth learning algorithm based on convolution neural network is adopted to support full convolution, full picture, and full tracking detection. The detection index is greatly improved and the detection function is more comprehensive. On the one hand, it can achieve illegal testing of pedestrians, pedestrians running red lights, illegally turning heads, turning left and not going straight, turning left and turning small corners. On the other hand, the driver can also achieve face image, driving phone calls, driving without seat belt detection; In addition, it can also support vehicle flow statistics and vehicle type, body color, brand logo and other vehicle attribute information recognition.