Face Fatigue Detection Method Based on MTCNN and Machine Vision

Abstract

Multi-task Cascaded Convolutional Networks (MTCNN) is a face detection method based on deep learning. Compared with the traditional parametric model and regression-based method, MTCNN is more robust to light, angle and facial expression changes in the natural environment, while machine vision as an important branch of the current artificial intelligence technology, it realizes the visual function of the human eye through a computer. By combining MTCNN with machine vision, real-time face detection can be realized. This detection method can be used for in the fields of identification, face behavior analysis, etc. it is especially suitable for the field of fatigue detection and early warning for long-term computer use. Based on this, this paper proposes a face fatigue detection method based on MTCNN and machine vision. This method combines three parameters of blink frequency, yawn frequency and drowsiness frequency, and uses fuzzy neural network to force computer users to know their fatigue in time. For office workers, it can improve their work efficiency and even prevent the occurrence of accidents. For ordinary users, it can remind them to protect their eyes and help them maintain body-health. With the promotion of this method, people can realize the danger of fatigue from the source and develop a healthy way of life and work.