应用区域估计的复杂背景多圆快速定位方法

来源期刊:中南大学学报(自然科学版)2016年第7期

论文作者:王宪 赵前程 谭建平

文章页码:2266 - 2273

关键词:计算机视觉;多圆检测;自适应滤波;收敛;区域估计

Key words:computer vision; multiple circle detection; adaptive filter; convergence; region estimation

摘    要:针对现有多圆定位方法应用于复杂背景高分辨率工业测量图像实时性差的问题,提出一种应用区域估计的复杂背景多圆快速定位方法。首先,依据Hough变换前2帧图像检测结果及环境噪声经验数据实现各圆特征对应的Sage-Husa滤波器的初始化;接着,应用滤波算法对特征的有效区域进行估计;最后,在有效区域内结合优化的特征定位有效判据及2种图像增强方法实现精确定位。研究结果表明:优化的特征定位有效判据较传统判据更加合理;特征提取速度与基于Hough变换的方法相比显著提高;应用该方法的测试系统可充分发挥视觉测量技术在灵活性及集成性上的优势,测试精度较传统系统提高33%以上。该方法可实现高分辨率图像多圆快速精确定位,具有实用价值。

Abstract: A multiple circle location method using region estimation in complex background was proposed in light of the poor real-time performance of existing circular positioning method applied in high-resolution industrial measurement image with complex background. Firstly, the initialization of the Sage-Husa filter for each circular feature was realized by the first two images testing results through Hough transform and the environmental noise empirical data. Then, the filtering algorithm was employed to estimate the effective area of the feature. At last, precise positioning combined the optimized positioning valid criterion and two image enhancement methods in the feature effective area were achieved. The results indicate that the optimized positioning valid criterion is more reasonable than traditional criterion, feature extraction speed is significantly improved compared with the method based on Hough transform and has better robustness. Moreover, test system employing this method can take full advantages of vision measurement technology in terms of flexibility and integration, and test precision is 33% greater than traditional systems. It can achieve fast and precise multiple circle location of the high-resolution image and has a practical value.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号