局部遮挡条件下的普通工件定位检测  

Ordinary workpiece positioning detection under partial occlusion

作　　者：余振军 孙林[1] 贾坤昊 孙洋 YU Zhen-jun;SUN Lin;JIA Kun-hao;SUN Yang(College of Geodesy and Geomatics,Shandong University of Science and Technology,Qingdao 266590,China)

机构地区：[1]山东科技大学测绘科学与工程学院,山东青岛266590

出　　处：《计算机工程与设计》2020年第10期2777-2783,共7页Computer Engineering and Design

基　　金：国家水体污染控制与治理科技重大专项基金项目(2017ZX07302-003)。

摘　　要：为解决工业应用中,普通工件在局部遮挡情况下难以精确定位的问题,提出局部遮挡条件下的普通工件定位检测算法。离线过程中,利用提取的模板图像轮廓角点特征和重采样后的轮廓点集,计算关键码值和形状描述子构建模板模型,建立几何哈希表(hash);在线过程中,计算实测关键码值在hash表中地址,利用投票机制对该地址下的形状描述子进行投票,实现目标匹配;基于匹配点对,基于Turkey权重函数和最小二乘平方差原理计算实测轮廓和模板轮廓间的仿射变换模型,实现目标精确定位。实验结果表明,当工件被局部遮挡,且发生平移、旋转变化时,可实现目标的精确定位。In the industrial application,ordinary workpieces are difficult to accurately position in the case of partial occlusion,to solve the problem,an algorithm of the ordinary workpiece positioning detection under partial occlusion was proposed.In the offline process,the corner feature and the resampled contour points set of template image were used to calculate key coded values and shape descriptors to establish geometric hash table.In the online process,the address of the measured key code value in the hash table was calculated,and the voting mechanism was used to vote on the shape descriptor under the address to achieve target matching.Based on matching point pairs,the Turkey weighted function and the least squares square difference principle were used to calculate the affine transformation parameters between the measured contour and the template contour to achieve precise positioning of the target.Experimental results show that the precise positioning of the target can be achieved when the workpiece is partially occluded and the translation and rotation change.

关 键 词：形状匹配 目标定位 多边形近似拟合 最小二乘 Turkey权重函数 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]