摘要
边缘特征是图像的最底层特征,解决好边缘检测问题对多数目标识别系统都有重要的意义.针对机场跑道的直线段特征,提出了一种有效的基于边缘图像的机场跑道识别算法.首先,将方向小波变换用于目标的边缘检测,得到定位准确、边缘完整、冗余信息少的边缘图;然后两次遍历该图,改进Hough变换在提取直线段的同时,获得直线段端点坐标,精确定位出跑道位置.实验结果表明了该算法的有效性和准确性.
Edge feature is the rock-bottomfeatures of image. Excellent edge detection is significant for most object recognition systems. Considered the linear feature of runway, an effective algorithm of runway recognition based on edge map was presented. Firstly, directional wavelet transform was adopted to obtain the edge map with ex- act localization, unabridged edges and less redundancy information; Then scanning the edge map twice, and ex- tracting the straight lines using the improved Hough Transform. Simultaneity, the start and end coordinates of the straight line segments were also acquired, which could localize the runway position accurately. Experimental results indicate that this algorithm is effective and precise.
出处
《光电技术应用》
2008年第1期70-73,81,共5页
Electro-Optic Technology Application
基金
重庆自然科学基金项目(CSTC
2005BB2207)
关键词
跑道识别
边缘检测
方向小波变换
HOUGH变换
直线段定位
runway recognition
edge detection
directional wavelet transform
Hough Transform
straight line localization
