摘要
为了满足对表面贴装器件(surface mounted devices,SMD)器件边缘高精度的定位要求,提出了一种改进的SMD图像亚像素边缘检测算法。首先,对SMD图像进行双边滤波和二值化处理;其次,采用引入信息熵加权系数的多类型多尺度的结构元素对预处理后的SMD图像进行形态学运算,获得像素级图像边缘;最后,在像素级图像边缘的基础上,采用一种改进模板系数和边缘判断条件的Zernike矩边缘模型对SMD图像重新定位,确定亚像素边缘坐标,实现SMD图像边缘检测。实验结果表明,该算法与Zernike矩和改进Sobel-Zernike矩算法相比,亚像素坐标与真实坐标误差更小且仅为0.4940 pixels,运行时间分别减少了54.671%和19.407%。
In order to meet the high-precision positioning requirements for surface mounted devices(SMD) device edges,an improved sub-pixel edge detection algorithm for SMD images is proposed.Firstly,the SMD image is processed by bilateral filtering and binarization.Secondly,the multi-type and multi-scale structure elements with weighted coefficient of information entropy are used to carry out morphological operation on the pre-processed SMD images to obtain pixel-level image edges.Finally,on the basis of pixel-level image edges,a Zernike-moment edge model with improved template coefficient and edge judgment conditions is used to reposition SMD images,determine sub-pixel edge coordinates,and realize SMD image edge detection.Experimental results show that compared with Zernike moment and improved Sobel-Zernike moment algorithm,the error between sub-pixel coordinate and real coordinate is less than 0.494 0 pixels,and the running time is reduced by 54.671% and 19.407% respectively.
作者
孙瑞彤
谷玉海
张英
Sun Ruitong;Gu Yuhai;Zhang Ying(School of Mechanical and Electrical Engineering,Beijing Information Science&Technology University,Beijing 100192,China;Key Laboratory of Modern Measurement&Control Technology,Ministry of Education,Beijing Information Science&Technology University,Beijing 100192,China;School of Automation,Beijing University of Posts and Telecommunications,Beijing 100876,China)
出处
《国外电子测量技术》
北大核心
2023年第9期130-136,共7页
Foreign Electronic Measurement Technology
基金
现代测控技术教育部重点实验室开放课题(KF20222223205)项目资助。
