| Since the beginning of the 21st century,with the rapid development of economy,people have a higher and higher demand for valuables.Automobile has become the first choice for people to consume.However,drivers' lack of awareness of driving,fatigue driving,drunk driving and driving in violation of traffic rules will lead to traffic accidents.According to the investigation and analysis of the causes of traffic accidents,fatigue driving is the main factor causing traffic accidents.In this study,the fatigue driving detection system at home and abroad was deeply studied,and the detection algorithms used in various detection systems were analyzed and compared to find a relatively simple fatigue detection algorithm.The main idea is to coarse position the human face,then fine position it to the human eye,and then judge whether the driver is in the fatigue driving state through the human eye state.First of all,this paper analyzes the dual-space color information and Adaboost classifier,and forms an effective face detection model.Specific process of this method is as follows:first the face rough location,the process of this method is based on skin color information needed in the different characteristics of HSV and YCbCr color space to complete the rough location of face;Secondly,the accurate detection of human eyes requires the construction of a new face classifier according to the Adaboost algorithm of harr-like features,and the high-precision positioning is completed based on the previously obtained rough areas.According to relevant experiments,this algorithm can improve the detection efficiency and ensure the accuracy and robustness of the Adaboost algorithm.Finally,this paper analyzes the factors influencing the driver's posture tilt and background complexity,and designs special improvement and adjustment measures.Secondly,in this paper,the human eye area template matching algorithm and pupil Hough transform circle detection algorithm has carried on the detailed study,based on this method can effectively detect eye position and state of opening and closing to the driver.Detect pupil circle is open eye state namely,it is closed eye state otherwise.Algorithm is designed in this paper also fully apply to the threshold segmentation method,the image of each frame to the human eye opening and closing state identification,according to the proportion of unit time eyes opening and closing,to determine whether the driver is in a state of fatigue driving.This algorithm can effectively reduce the fatigue detection time of drivers and improve the calculation efficiency and accuracy of the algorithm.In terms of fatigue detection,this algorithm will have greater advantages.Finally,the paper completed the algorithm based on C language programming,and compile the application of the algorithm,but also using the OpenCV library to establish a set of efficient fatigue detection system,through a series of optimization and improvement makes further enhance the effectiveness of the algorithm.Nowadays,with the frequent traffic accidents,the application of intelligent automatic fatigue driving detection technology is of great significance to the prevention of traffic accidents. |
