Application Of Gabor Multiscale Fusion Feature In Fatigue Driving Detection

With the increase of the number of cars and the acceleration of the pace of people’s life,traffic accidents caused by driver fatigue occur frequently.It has become a hot and difficult point in the field of fatigue driving detection technology to predict the driver’s fatigue state in advance so as to avoid traffic accidents.Facial expression recognition and its analysis is an attractive research in the field of computer vision application and pattern recognition.Facial expression is more abundant than any other part.When human beings express the corresponding feelings,the facial expression changes immediately.The face makes micro expression through muscles.The areas that express the most emotion are the eyes,mouth,and cheek.When the face is tired,the eyes are slightly closed,the eyes are mindless,the eyebrows are downward,the cheek muscles are relaxed,and the mouth is slightly closed.The above characteristics can be used to judge whether the driver is in fatigue state.(1)A fatigue recognition method based on Gabor feature fusion of facial expression is proposed.In order to fully capture facial emotion information,this paper uses Gabor multi-directional feature fusion and block histogram method to extract facial features from the region of interest.Gabor filtering has a good expression effect on the face space.The fusion features can effectively reduce the redundancy between the original Gabor feature data,and can also the multi-scale analysis of the region of interest is carried out,and the block histogram strengthens the description of global features.The above methods ensure that the decision-making information will not be lost and the redundant data will be eliminated,which increases the processing efficiency and accuracy.(2)A new comprehensive fatigue determination strategy has been formulated,and drivers will have different physiological mechanism responses according to mild fatigue,moderate fatigue,and severe fatigue during driving.When the driver is in different degrees of fatigue,the brain will produce corresponding resistance and enter a new cycle of excitement.The system makes fatigue judgments based on the other basic expressions of the driver during the driving process and the proportion of fatigue expressions.This strategy combined with fatigue expression recognition can effectively prevent fatigue driving.(3)Fatigue warning software based on facial expression is designed.An image input module,a video image preprocessing module,an image feature extraction module,an image expression classification module,and a fatigue analysis module are designed.Complete the fatigue analysis and detection function of driving video simulation through the above modules.This system is designed and implemented through the Py Charm platform using Python language.