Development Of Infrared Pattern Projection-based Experiment System For Detecting Driver’s Head Posture

Real-time detection of the person’s head posture has become more and more important in the fields of safe driving of automobile, man-machine interface of game and robot. At present many domestic and foreign research groups have devoted a lot to the detection of the person’s head posture. However, in these investigations two dimensional image processing methods are used and the person’s head posture is determined by analyzing the relative position relations of facial features, the limitation of which is that the results are strongly influenced by ambient light, and can estimate the face directions within a small angular rangeIn order to solve these problems, in this thesis we proposed a infrared pattern projection-based method of estimating the head posture of the driver by three-dimensional (3D) image analysis. First of all, steps of whole experiment system and the basic principles of3D measurement method were introduced, and then each key step of the image processing was described in detail. Second, based on experiment conditions of both hardware and software, every aspect of the experimental system was considered to complete the development of the whole system. Finally, using mannequin and real people as the research object, the measurements were performed by using3-D image analysis which is based on the infrared pattern projection technique. And preliminary analysis of the experimental results was discussed. The work is helpful for the future development of a practical detection system.