The Research On Monitoring System Of Image Preprocessing Technology Of Safe Driving

With the development of transportation and the increasing speed, traffic accidents have become the current serious problems for all countries. Meanwhile, the number of traffic accidents due to driver fatigue is increasing largely, traffic accidents caused by driver fatigue have become one of the main factors. Therefore, how to monitor and prevent fatigue driving effectively is an important practical significance to reduce traffic accidents and mortality. With the development of Image Processing Technology, the real-time detection systems for driver fatigue based on computer vision has bright prospects, especially in high quality active automobile safety and accident avoidance measurement. In the using of computer vision to monitor the driving behavior, image preprocessing is a crucial section.As to degraded image acquisition from driver fatigue monitoring system, Image pre-processing algorithm proposed mainly includes several sections as follows:Firstly, according to the image to a dilemma in the problem of how to reduce the noise in theimage noise at the same time, as much as possible to retain the image details, proposes the denoising method for infrared image based on Curvelet transform. Since the representation of wavelet transform to the optimal high dimensional function containing line singular or singular,the existence of which makes the wavelet transform in the analysis of two-dimensional imagecurve or straight line edge feature is obviously insufficient, used in the infrared images to nobetter approximation accuracy and noise in sparse expression ability. For this, put forward a kind of improved Curvelet transform based threshold denoising algorithm, using the soft threshold and hard threshold combination, form the new threshold function. Secondly,according to the cab daylight illumination caused by uneven image is not clear, adopts the method of wavelet transform based on image contrast inhomogeneous has obviouselimination, improves the contrast of the image and quality, by using the method of wavelet transformation image after correction, not only can eliminate the uneven illumination, but alsohas the smallest error correction method based on wavelet transform has less time consumption. Finally, according to the cab caused by brightness image contrast weak problem, adopt adaptive enhancement algorithm of infrared image based on Gauss mixture model, histogram statistics using face image to the collection of the best state, the provisions of the pending gray histogram distribution image face part, so as to enhance the image contrast, enlarging gray range image, the details of the image more clearly. Respectively to the standard and real images experiments are carried out. And the experimental results are given for each algorithm, the new proposed algorithm and the existing algorithms were compared, and analyzed the advantages and disadvantages of the algorithm compared with existing algorithms proposed new. Experiments show that, the new algorithm has much more superiority in the image preprocessing.