基于双目视觉的前车距离识别方法

本文提出一种基于双目视觉系统,对道路上各对象进行特征识别的方法。首先利用Canny算子和霍夫变换等图像处理方法检测出道路区域,以提高其后的图像处理效率;然后在道路区域内部通过阴影识别方法对车辆进行识别;最后利用双目视觉几何关系,对对象物体距离进行比较。实验结果表明,通过本文所述方法可准确识别出图像中各对象及其距离关系。

Distance Detection Method of Forward Vehicle Based on Binocular Vision System

This paper describes an on-road objects detection method which is based on a binocular vision system. At first, for improving the image processing speed, canny algorithms and Hough transform are used to recognize road area. Then, vehicle （：an be distinguished by shadow extension. At last,the distance relations of each object in the image will be ob rained. The result shows that all main objects and distance relations of them can be detected accurately.