改进的Hough变换直线段检测方法

The Improved Approach of Straight Line Segments Detection Based on Hough Transform

Hough变换是一种很有效的检测直线段的方法,然而传统的Hough变换法检测直线存在运算量大、断线错连等缺陷。本文通过分析Hough变换原理,提出利用梯度阈值筛选有效像素点,用间隔阈值分割直线的改进方法。经实验证实,降低了计算复杂度并避免了小直线段的丢失。

The Hough Transform is an efficient algorithm in straight line segments detection field, however, there are deficiencies in the classical Hough transform, such as the great deal of operation, the wrong connection of the break-line and etc. By analyzing the principle of the Hough transform, we improve a method to make a threshold to avoid inefficient pixels, and another threshold to segment straight lines. It is proved that, compared with classical Hough transform, this method reduce the complexity of operation, and avoid the lost of short straight lines.