Fatigue driving detection method and system based on facial feature fusion

Abstract

The invention relates to a fatigue driving detection method and system based on facial feature fusion, and belongs to the technical field of fatigue driving detection. The method comprises the following steps: firstly, carrying out face region positioning on a frame of image in acquired image information through a high-precision face detection algorithm, ensuring the accuracy of face position recognition; dynamically tracking the rest frames of images by taking the positioned face area as a target; obtaining a face area of each frame of image of the driver. On the basis that a high-precision face detection algorithm is not continuously adopted, face position recognition of other frames of images is achieved through a dynamic tracking algorithm with less time consumption, the face positioncan be accurately detected, the judgment accuracy of fatigue driving is improved, and the real-time performance of the detection algorithm can be guaranteed.

Description

technical field [0001] The invention relates to a fatigue driving detection method and system based on facial feature fusion, belonging to the technical field of fatigue driving detection. Background technique [0002] At present, the commonly used fatigue driving detection methods mainly include methods based on physiological information, methods based on computer vision and methods based on changes in vehicle driving behavior. The method based on physiological information is the method with the highest accuracy rate in fatigue driving. Physiological signals such as electroencephalogram (EEG) / electrocardiogram (ECG) / electromyogram (EMG) can reflect the driver's condition in a timely and accurate manner. Fatigue degree, the disadvantage is that the measuring instrument is expensive and needs to be worn by the driver on the body, which interferes with normal driving and affects comfort; the method based on vehicle driving behavior is non-invasive and does not interfere with t...

AI Technical Summary

Problems solved by technology

Although this method can realize the positioning of the human face, Adaboost cannot accurately locate the position of the human face when the driver wears glasses, sun glasses, light changes, and partial facial occlusion, and thus cannot accurately judge fatigue driving.

For this reason, some people propose to use the MTCNN algorithm with higher precision for face detection. Although the MTCNN face detection algorithm can accurately detect the position of the face in the presence of the above-mentioned interference, the detection method has relatively high hardware requirements. High, and the method is complex, time-consuming, difficult to meet real-time driving requirements

Images