一种改进随机Hough变换的圆形检测算法  

An improved circular detection algorithm based on Randomized Hough Transform

作　　者：李建森 LI Jiansen(Huangshan University,Huangshan 245041 China)

机构地区：[1]黄山学院信息工程学院,安徽黄山245041

出　　处：《新余学院学报》2025年第2期41-47,共7页Journal of Xinyu University

基　　金：安徽省教育厅自然科学重点项目“基于稀疏数据的弹性关系模型研究——以村落元数据价值增强为例”(2023AH051371);黄山学院自然科学研究项目“大数据时代古村落深度游的多维度研究”(2020xzx004);国家级大学生创新创业训练计划项目“基于深度学习的复杂场景下交通标志识别研究”(202310375055)。

摘　　要：针对随机Hough变换存在占用空间大、计算量大等缺点,提出一种改进随机Hough变换的圆形检测算法。对原始图像进行图像增强、边缘检测等处理得到边缘图像,以减少边缘点的数量;在边缘图像上随机选择一边缘点,利用圆形特征找到该点在横纵方向上的另两点来确定候选圆,降低无效采样的次数;利用三级筛选法统计候选圆特殊方向上的边缘点来判断候选圆的真假。实验检测结果表明,该算法相比RHT算法和RHT_S算法,在效率和准确率上均有提高,且可靠性更高。Aiming at the shortcomings of Randomized Hough Transform such as large space occupation and large calculation amount,an improved circular detection algorithm based on the Randomized Hough Transform was proposed.Firstly,the original image was preprocessed,and the edge image was obtained through image enhancement and edge detection to reduce the number of edge points;Then an edge point was randomly selected on the edge image,and the circular features were used to find the other two points of the point in the horizontal and vertical directions to determine the candidate circles,reducing the number of invalid samples;Finally,a three-level screening method was used to count the edge points in the special direction of the candidate circle to determine the true or false of the candidate circle.The experimental results show that compared to the Randomized Hough Transform algorithm and RHT_S algorithm,this algorithm has improved efficiency and accuracy,and has higher reliability.

关 键 词：随机HOUGH变换 圆形检测 圆形特征 三级筛选 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]