随着近几年图像处理技术以及计算机处理能力的飞速发展,视觉导航已成为机器人导航的主要发展方向之一.在此采用彩色空间分割方法,改进了基于区域直方图的分割算法,获取道路YUV空间上下阚值,根据阈值在三维空间中进行全图分割得到二值化图像.并提出了一种新的分段扫描式路径检测方法,对摄像机进行分段标定,给出可行走线区域及其端点的世界坐标值,实现了路径的快速、准确检测.%By the rapid development of the image processing technology and computer processing function in recent years, the vision navigation has already become one of the main developing directions of robot navigation. The segmentation algorithm based on region histogram was improved with the method of color space segmentation to acquire the up and down threshold of the road in YUV space, and obtain two-valued image by full image segmentation in 3-D space according to these thresholds. A new method of subsection route scanning detection are brought forward to execute the subsection calibration of camera. The areas of walking line and its extreme point's coordinate in the word model are offered. Therefore, the rapid and correct dectection of road was implemented.
展开▼
