The invention discloses a driver fatigue detection method and device based on a convolutional neural network, belonging to the field of artificial intelligence, and aims to solve the technical problemof how to detect the fatigue degree of a driver by using the convolutional neural network and give an alarm to the fatigue driver, and the technical scheme comprises the following steps: S1, collecting face image data under various conditions, and carrying out feature marking on images; S2, training the image data set by adopting a cnn convolutional neural network algorithm to obtain a corresponding model; sS, compiling a corresponding python code loading program for the model trained in the step S2; S4, installing the loading program in the step S3 into a host of a center console in a cab; and S5, enabling the vehicle-mounted camera in the cab to shoot pictures of the driver and transmit the pictures to the loading program for reasoning, and judging whether the driver is in fatigue driving or not through the trained model. The invention further discloses a driver fatigue detection device based on the convolutional neural network.