一种基于精简粒子群优化的霍夫变换算法.pdf

第 44 卷 第 2 期 天 津 大 学 学 报 Vol.44 No.2 2011 年 2 月 Journal of Tianjin University Feb. 2011 ▋ 一种基于精简粒子群优化的霍夫变换算法 张英涛，黄剑华，唐降龙，刘家锋 (哈尔滨工业大学计算机科学与技术学院，哈尔滨 150001) 摘 要 ：为了提高现有霍夫变换算法的准确率及计算效率，提出了一种基于精简粒子群优化的霍夫变换算法．该方法 将霍夫变换后解的参数作为粒子位置，将霍夫变换的累加数组值作为粒子的适应度值，每一次迭代更新粒子的位置和 速度，并将所得粒子群的适应度值按降序排列，保留“强壮”粒子，组成精简粒子群．实验结果表明，基于精简粒子群 优化的霍夫变换算法不仅可以提高霍夫变换的计算速度，而且可以获得较高的计算准确率，特别是对于复杂背景图像 和高噪声图像也同样适用． 关键词 ：霍夫变换；精简粒子群优化；图像处理；曲线检测 中图分类号：TP391.4 文献标志码 ：A 文章编号 ：0493-2137 (2011)02-0162-06 A Hough Transform Algorithm Based on Reduced Particle Swarm Optimization ZHANG Ying-tao ，HUANG Jian-hua ，TANG Xiang-long ，LIU Jia-feng (School of Computer Science and Technology ，Harbin Institute of Technology ，Harbin 150001 ，China) To improve the accuracy and efficiency of existing Hough transform algorithms ，a novel Hough transfor- Abstract ： mation algorithm based on reduced particle swarm optimization (RPSO) is presented. The parameters of the solution after Hough transformation are employed as particle positions ，and an accumulation array in Hough transformation is utilized as a fitness function of RPSO algorithm. Velocities and positions are updated accordingly ，and position fit- ness values are calculated and sorted in a list with the descending order. “Stronger” particles whose fitness values are in the higher positions of the list are selected and the reduced particle swarm is then obtained. The experimental results show that the proposed approach can detect curves more