Research Of Road Detection Algorithm Based Of Monocular Vision

Recent years, Intelligent Vehicle has become a new field which attracts more attention over the world. Because of ts unique advantage, the Vision Navigation system takes a special place in this field. One nice Intelligent Vehicle System, should detect the real environment in case of driving and locate the boundary of the road in real-time so as to guarantee the intelligent vehicle to drive autonomously steadily and safely without manual operation.For structural road, this paper focuses on the detection of lane line. Firstly researching with the classic algorithm about the image preprocessing and image edge detection, then this paper address that use Median of the Intercepts and Catmull-Rom splin function to extracts road dividers line. This algorithm is focus on the straight and turning road. The test result are given, which show that the algorithm is efficient in white lane under all kinds of illumination environment, and it can estimating the direction of road extension accurately.This paper proposes the concept which based on the Semantic Mapping method to detection unstructured road. Use the wild unstructured semantic model to change all reality boils down to be one model, that makes this system less calculation and higher speed, and satisfies robust real-time basically. It also resists the effect of shadows and water marks.Finally, this is fist introduction of Random Forest to complete road-detection. Applying with Random Forest for the classification of searching a best classification algorithm, can promote the precision of the detection.