Application of Traffic Image Recognition in Snow Driving Safety

摘要

In order to improve the safety of vehicles in snow, this paper mainly studies traffic image recognition in snowy weather. Firstly, based on previous research experience, the formation mechanism of snowy weather is briefly described. On this basis, the influence of snowfall on road traffic and its characteristics are analyzed. Then, the real-time snowfall level image acquisition and recognition is carried out. Different from the traditional method of dividing the snow level according to a 24-h snowfall, this paper distinguishes the real-time snow level according to the difference of the gray value of the image caused by the influence of snow on image clarity. Finally, according to the influencing factors of road traffic safety in snowy weather, combined with the result of image recognition, an early warning system for driving in snowy days is constructed.