基于改进Zernike矩的玻璃瓶亚像素边缘检测算法  

Improved Zernike Moment-Based Method for Subpixel Edge Detection of the Glass Bottle

作　　者：杨浩[1] 裴蕾[1] 李昌顺[1] 杨梅[1] 

机构地区：[1]重庆大学输配电装备及系统安全与新技术国家重点实验室,重庆沙坪坝区400044

出　　处：《电子科技大学学报》2012年第6期875-880,共6页Journal of University of Electronic Science and Technology of China

基　　金：输配电装备安全国家重点实验室项目(1010102020090116)

摘　　要：图像的边缘检测精度决定了实际尺寸的测量精度,为了提高测量精度,提出了一种基于改进的Zernike矩的测量玻璃瓶实际尺寸的算法。将传统的Zernnike矩算法与Otsu自适应阈值法相结合,得到改进的快速算法;利用该快速算法对采集的玻璃瓶图像进行亚像素级边缘检测;运用最小二乘法根据亚像素坐标对瓶口与瓶底所形成的椭圆进行拟合;应用标准块规对系统进行标定并得到玻璃瓶实际的测量尺寸。实验结果表明,该方法不仅可使边缘检测达到亚像素级精度,还避免了人工反复选择阈值而造成的低效率与误判,实现了快速、准确、无接触测量实际尺寸的功能。As the accuracy of edge detection determines the accuracy of actual size measurement, in order to improve the measuring accuracy, this paper proposes that a faster algorithm should be adopted to detect the glass boule dimension based on the Zernike moments. The traditional Zernike algorithm is combined with Otsu adaptive threshold algorithm to get a faster improved algorithm. The edge of glass boule with subpixel level is detected by using the improved algorithm. The ellipses formed by the glass bottle＇s mouth and bottom are fitted by applying the least square method. At last, the system is calibrated by gauge block to get the actual measuring dimension. Experimental results show that the improved algorithm can not only make the edge detection reach the subpixel accuracy, but also avoid the edge misidentification and inefficient which will be caused by repeatedly manual adjustments to select the threshold value when we detect the edge.

关 键 词：标定 最小二乘拟合 Otsu自适应阈值法 亚像素 ZEMIKE矩 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]