显微视觉系统对柱状微零件自动聚焦技术研究.docxVIP

1222014，50（8）ComputerEngineeringandApplications计算机工程与应用

显微视觉系统对柱状微零件自动聚焦技术研究

罗李焱，徐德，张正涛，张娟

LUOLiyan,XUDe,ZHANGZhengtao,ZHANGJuan

中国科学院自动化研究所精密感知与控制中心，北京100190

ResearchCenterofPrecisionSensingandControl,InstituteofAutomation,ChineseAcademyofSciences,Beijing100190,China

LUOLiyan,XUDe,ZHANGZhengtao,etal.Autofocusmethodofmicroscopevisionsystemforcolumnarmicro-parts.ComputerEngineeringandApplications,2014,50（8）：122-126.

Abstract：Onaccountofcolumnarpartsinmicroscopevisionsystem,anautofocusmethodofthereal-timeimagecharac-teristicsdetectingandfeatureareatrackingispresented.Focusevaluationfunction,focussearchingalgorithmandfocus areaselectingisincludedintheproposedfocusingalgorithm.Thesystemcanachieveautofocusquicklyandaccurately,andtheshortcomingsofhill-climbingsearchingalgorithmandlocalmaximumofthesearchingresultcanbeovercome.Meanwhile,itisimportanttoselectanappropriatefocusarea.Theimagefeatureareaisusedasfocusareainthepaper.Themicroscopevisionsystemfocusesafterdetectingtheimagefeature.Itcanfocusaccuratelyatthecolumnarparts’edgeinthesituationsoftheradiusofthecolumnarpartsisshorterthanthedepthofscope.Thestandardvarianceerroris 205μmforcoarsefocus,and37μmforfinefocus.Experimentalresultsprovetheautofocusmethodcansatisfytherequestofthemicro-assemblysystemforfocus.

Keywords：focusareaselecting;focussearchingalgorithm;autofocus;columnarparts;micro-vision;micro-assembly

摘要：针对显微视觉中的柱状物体图像的清晰度问题，提出了一种实时检测图像特征并跟踪特征区域进行自动聚焦的方法。该聚焦算法包括聚焦评价函数、聚焦搜索算法和聚焦区域选择。该聚焦搜索算法实现了显微视觉系统下快速准确的聚焦，克服了爬山搜索算法的缺点，有效避免搜索结果陷入局部极大值。显微视觉中对聚焦区域的选择尤为重要，以图像特征区域作为聚焦区域，实时检测该特征区域进行聚焦，实现了在景深小于柱状物体半径的情形下对柱状物体边缘的精确聚焦。粗聚焦后电机位置的标准差为205μm，精聚焦后电机位置的标准差为37μm。实验结果表明，该自动聚焦方法能够满足微装配系统对显微视觉的聚焦需求。

关键词：聚焦区域选择；聚焦搜索算法；自动聚焦；柱状零件；显微视觉；微装配

文献标志码：A中图分类号：TP393doi：