Fatigue state detection method and system based on key point detection and head posture

Abstract

The invention discloses a fatigue state detection method and system based on key point detection and head attitude, and the method comprises the steps: constructing and training a backbone network, employing an MMC multi-task prediction model of a depth separable convolutional network, obtaining a plurality of frames of face images in unit time, employing an MTCNN network to detect the face position of each image, and cutting out a head image; inputting the head image into a trained MMC multi-task prediction model to obtain position information of a head posture angle and a face key point; respectively judging the fatigue states of the head, the eyes and the mouth by using a double-threshold method; the method comprises the following steps: setting a correlation coefficient to comprehensively judge the fatigue state of a person, combining the correlation of face key point detection and head posture, adopting an MMC multi-task prediction model with a backbone network as a deep separable convolutional network, and simultaneously performing two tasks in the same network, thereby greatly reducing the required parameter quantity and operation quantity, and improving the accuracy of the prediction result. Therefore, the detection speed of the model is improved, and the real-time effect is achieved.

Description

technical field [0001] The invention belongs to the technical field of automatic fatigue detection, in particular to a fatigue state detection method and system based on key point detection and head posture. Background technique [0002] Face detection and head pose estimation in the field of computer vision refers to detecting all faces in an image and estimating that each face can represent three orientation angles: yaw (yaw), pitch (pitch) and roll (roll) . Judging people's motives and intentions based on head posture has a wide range of applications in providing cues and gaze, such as human behavior analysis and gaze estimation. Although face detection and pose estimation have made great progress respectively, it is still a daunting task to realize a multi-task framework with good real-time and robustness in complex environments. At present, the convolutional neural network (CNN) is generally used to solve face detection and head pose estimation, which can achieve extr...

AI Technical Summary

Problems solved by technology

At the same time, the common convolutional neural network model is relatively large, and it is difficult to achieve real-time results when predicting

Images