一种高效的Hough变换实现方法

An Efficient Implementation of the Hough Transform

在图象处理和计算机视觉中,Hough变换(HT)是一种应用非常广泛的形状分析技术,但HT的一个主要缺点是它需要大量的计算时间.本文提出了一个用查找表实现的高效的HT方法,用于对直线和圆的检测。实验结果表明,新算法的计算速度比经典的HT算法约提高了一个数量级。

The Hough transform, HT, is a technique used very widely for shape analysis in image processing and computer vision. However, the principal disadvantage of the HT is its large computational requirement. In this paper, we propose an efficient table lookup implementation of the HT. The proposed algorithms are used for detecting lines and circles in pictures. The experimental results show that our algorithms are about one order of magnitude faster than the conventional HT algorithms.