Visibility Detection Technology Based-on High-Definition Video

The fog, rain or other bad weathers and insufficient illumination may lead to unclear scenery, vision with poor quality and low contrast. Fog and other bad weather will bring great harm to the transportation, production and people’s health. One of the greatest harm is the reduction of visibility and then traffic congestion and delays, even a large number of traffic accidents associated will occur, especially on the highway.The accidents caused huge losses and did serious harm to people’s lives and property. According to statistics, the number of traffic accidents due to fog is up to40%in a year. The number of highway accidents caused by fog is accounted for more than1/4of the total number of accidents. And the accident rates are increasing every year.In view of the adverse weather conditions, especially the significant impact of road traffic safety caused by fog, it is the important problem for experts, engineering and technical staff in related fields committed to tackling that combining advanced information technology, data communications transmission technology and electronic control technology, computer processing technology to protect the road driving security in the fog conditions. From Koschnfieder law and other related theories formed up to now, the research on visibility detection instruments did by domestic and foreign researchers have been conducted for many years. The researches have achieved great results, all of which has its own advantages. The instrumentshave beenapplied to different areas and sectors. But these instruments or methods have a certain lack.In this paper, drawing on the basis of existing research results, a new visibility detection algorithm——visibility detection algorithm based on high-definition video is posed.First, lane as reference, the parameters of the camera can be calibrated by camera calibration technique and then the distance information of road is in image solved, that is, the depth information. After that, the atmospheric transmission——t can be obtain through the Dark Channel Prior method and optimized by soft matting technology. Using the depth information and the transmission having obtained, the atmospheric extinction coefficient can be solved. And then the current visibility value can be determined according to the visibility detection principle. This algorithm combines digital image processing technology, computer technology and others. Throughout method, the detection results of visibility distance are consistent with the results of instrumental method. The existing traffic monitoring systems and surveillance video are full used, so as to provide more effectively protect for traffic safety, especially highway traffic safety.