一种基于梯度Hough变换的同心圆检测新方法  

A New Gradient Concentric Circles Detection Method Based on Hough Transform

作　　者：陈星[1] 

机构地区：[1]厦门大学自动化系,福建厦门361005

出　　处：《自动化博览》2011年第7期86-89,共4页Automation Panorama1

摘　　要：目前检测圆的方法很多,最常用的为Hough变换,另外还有一些改进的Hough变换和圆的快速检测方法,但是这些方法如果直接运用于存在同心圆的图像中,就不能检测出同心圆。目前对同心圆检测的算法较少,而且存在一定局限性,这里提出一种结合圆梯度信息和二次检测圆的新的检测同心圆的算法,改善了检测的局限性,不需要事先确定圆的一些参数仍然能准确检测同心圆,具有一定抗干扰性,且检测速度较快。本文通过实验仿真和应用实例,证明了该算法简单、准确,有效。There are several methods to detect circles and Hough transformation is the mainstream method. Besides, Hough transformation has been improved and many rapid detection algorithms have been developed. However, these algorithms cannot detect concentric circles if applied directly. Existing concentric circles detection algorithms are rare and with certain limitations. This paper proposes a novel concentric circles detection algorithm based on circle gradients and secondary circle detection, which lowers the detection limitation and complexity. This algorithm is to some degree robust to the noise. It can detect concentric circles accurately without requiring apriori knowledge about the circles. Experiments and some applied examples in this paper show that this algorithm is accurate, quick and efficient with detecting concentric circles.

关 键 词：HOUGH变换 圆 同心圆 二次检测 梯度 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]