In recent years, traffic accidents have occurred frequently, and fatigue driving has been one of the main causes of traffic accidents, causing huge casualties and property losses. Aiming at the problem of fatigue driving, this paper designs and implements a driver driving safety monitoring system based on Android system. The system consists of a smart car instrument with an Android system and an external camera. Firstly, the facial video of the driver during driving is obtained through the external Android camera. Through the SeetaFace face recognition engine system and the two-level cascading MobileNet-V2 training model, the face detection and facial key point coordinate extraction are realized. Finally, the fatigue algorithm based on face key points is used to realize real-time monitoring of driver fatigue. Experiments were carried out in the case of a simulated cab. The experimental results show that the system is robust to light illumination and glasses, and can greatly reduce the incidence of accidents and ensure driver′s driving. Safety.