Research on Face Recognition Early Warning in Dangerous Areas Based on Raspberry Pi

A study on face recognition and early warning in dangerous areas based on the Raspberry Pi is proposed. It combines the Raspberry Pi with the improved “Multi-Scale LBP + CNN” algorithm to search for eigenvalues, extract features accurately and efficiently, and improve the accuracy of face recognition. At the same time, it can monitor human bodies and objects and issue voice warning reminders. The experimental results show that the latency times for recognition at different distances and from the front and side are different, but the recognition accuracy is very high.