摘要： 提出了一种基于Hough变换的前视机场跑道识别算法。该算法基于跑道灰度图的直线边缘特征,利用直线边缘特有的灰度方向和直线方向相一致的特点改进了基于二值化边缘点的传统Hough法,并在此基础上实现了复杂背景下的机场跑道识别。真实跑道图像测试表明本算法具有良好的识别效果。

关键词: 计算机视觉, 跑道, Hough变换, 边缘提取

Abstract: A recognition algorithm for forward-looking runway is presented based on improved Hough transform.The forward-looking runway is defined as a pair of linear edges,which are unifications of both the gradient direction and the line direction. With this definition the traditional Hough transform based on binary edge image is improved in order to extract the straight runway lanes form complex background. It is demonstrated to be a practical approach to runway recognition for the good simulation results on real images.

中图分类号: 

• TP391.4