Research Of Driver's Fatigue State Detection Technology Based On DM6437

Currently, driving fatigue has been a significant factor of traffic accident. Statistics show that traffic accidents also increased rapidly, it became a threat to people's life security as well as social order and stability. In order to protect the traffics security and to prevent traffic accidents, it is very necessary and meaningful to research effective ways to detect driver fatigue status and fatigue alarm.An improved algorithm is proposed based on the PERCLOS (Percentage of Eyelid Closure Over the Pupil Over Time, referred to as PERCLOS) method, the human face location is taken as the main detection target, and several simple and efficient image processing and mode recognition algorithms are combined to realize it. The main work is as follows:Firstly, the ICETEK-DM6437-B evaluate module is introduced simply, the software architecture and application development process of video loopback are analyzed in detail, which in order to processing effectively and in real-time on DM6437 DaVinci processor.Secondly, two face detection methods are proposed which based on color image and infrared gray image respectively. When considering the real-time requirements of system, images are diminished firstly, and then the driver's face region are segmented from background based on skin color segmentation model of YCbCr color space and binary segmentation model of infrared gray image respectively, image scaling method based on the two face segmentation mode with some other image processing algorithms are used to detect and locate the driver's face region.Finally, on the basis of face region detection and location face area, the current video frame is detected through the face area threshold value whether is fatigue frame or not. It statistics about the total number of fatigue frame, and then calculates the value of PER-NOFACE to determine the status of the driver.