Study on the detection of locomotive driver fatigue based on image

In order to overcome the traditional shortcomings of the contact-type fatigue detection for locomotive drivers, this paper presents a system that detects face in images collected by the CCD, then establishes the eye´s position in the region of the detected face, to determine the degree of the driver´s fatigue through the analysis of the state of the human eye. In this paper, AdaBoost method based on Haar features is applied to detect the face, on the basis of which, through the four templates matching method, to ultimately identify the position and status of the human eyes. Experiments show that, AdaBoost algorithm combined with the template matching method can be carried out with high detection accuracy and high speed, meanwhile, has good robustness for the changes of environmental conditions, and can meet the requirements of real time and accuracy in the fatigue detection.