Research On Road Detection Algorithms For Expressway

Vision-based navigation technology is a new hot spot in the area of autonomous navigation vehicle research. Images describe roundly the environment, and carry a tremendous amount of information about the scene. This article mainly discusses the road detection technique in a variety of highway conditions based on monocular vision system.During vehicles running on the road, objects'reflection and other complexity conditions always appeared on the road and would process an unsatisfactory result for road detection. So,a method which combined color segmentation with edge detection was presented in this paper for resolving those problems. The accuracy of color segmentation and the rapidity of edge detection were combined in this method, the experimental results showed that the method had feasibility and effectiveness.The traditional road recognition algorithms had their own disadvantages.In order to resolving those disadvantages, an improved method based on cluster analysis and region growing for road detection was presented after analyzing and comparing many traditional methods of road detection. The most important idea about the improved method was an adaptive region growing criteria. According to the distribution of characteristics of the image, the image was divided into three pieces and in order to reducing the processing time, each of pieces was processed by different method. In order to improving the accuracy of road detection, the white lanes of road were pre-processed into the same color as the road color reference value for eliminating interferences of road lane. The experimental results showed that this improved method had advantages of fast, high accuracy, reliability and robustness by processing many videos which size was 390×288 pixels.The accuracy on the road detection could be maintained at 95% and the processing time for each frame could be controlled within 0.4s.