对导航标识线边缘特征的提取进行了研究,具体研究了标识线特征提取的算法,并利用VC++对算法进行了实验,验证了该方法可以有效的进行边缘特征提取.针对基于视觉导航的AGV导航标识线的边缘特征提取,提出了一种新颖的边缘提取方法.首先用迭代阈值选择法对图像进行阈值分割,生成二值化图像,再用一种改进的中值滤波方法对图像进行平滑处理,最后用边缘追踪法对平滑后的二值图像进行边缘提取.并通过VC++使用所提出的算法对标识线图像进行了处理,成功提取出了清晰的边缘图像.%It researched the edge feature extraction of the navigation identification line, and specified the study of the line feature extraction algorithm,and used VC++ to take an experiment by the algorithm, showing that the method can extract edge feature effectively. For extracting edge feature of visual navigation marking line of AGV based on visual navigation, it presented a novel edge extracting method. Firstly, it used iterative threshold selection method to segment the image,and generate the binary image ,then it used an improved median filter on the image smoothing. Finally it used edge tracing method to extract edge of the smoothed binary image. And it used the proposed algorithm to process the marking line image in VC++, successfully extracting a clear image of the edge.
展开▼
