Research Of The Road Obstacle Detection

Obstacle detection in the intelligent vehicle vision navigation system occupies a very important role. This article is based on computer vision technology to detect obstacles on the process of research. The study includes three areas: Camera imaging and the design of monocular vision model, obstacle detection and obstacle distance measurement.In obstacle detection, we detect the obstacles on the base of the road region. First of all, use the median for filtering, use Sobel operator to detect obstacles on the edge, after, get the basic information about obstacles by using the morphological processing, and then scan the image line by line, finally, mark the obstacles by a rectangular box.To measure distance of obstacle in front, this article will simplify monocular vision system to the camera projection model. Establish the model of monocular measurement through the geometric relationship, obtain the transformation relationship between image coordinate and world coordinate system, and finally achieve the obstacles of distance measurement.