测绘学报 ›› 2015, Vol. 44 ›› Issue (1): 46-51.doi: 10.11947/j.AGCS.2015.20130795

• 摄影测量学与遥感 • 上一篇    下一篇

附约束条件的零件轮廓线的多特征提取

郭宝云1, 李彩林1, 黄荣永2   

  1. 1. 山东理工大学建筑工程学院, 山东 淄博 255049;
    2. 武汉大学遥感信息工程学院, 湖北 武汉 430079
  • 收稿日期:2013-12-24 修回日期:2014-06-04 出版日期:2015-01-20 发布日期:2015-01-22
  • 通讯作者: 李彩林 E-mail:licailin@whu.edu.cn
  • 作者简介:郭宝云(1986-), 女, 博士, 讲师, 研究方向为近景摄影测量、计算机视觉等. E-mail: guobaoyun@sdut.edu.cn
  • 基金资助:

    国家自然科学基金(41171357);山东理工大学博士科研基金(4041-413050;4041-413042);中央高校基本科研业务费专项资金(2012213020212)

Multiple Features Extraction of Part Contour with Restrictive Constraints

GUO Baoyun1, LI Cailin1, HUANG Rongyong2   

  1. 1. Institute of Architectural Engineering, Shandong University of Technology, Zibo 255049, China;
    2. School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China
  • Received:2013-12-24 Revised:2014-06-04 Online:2015-01-20 Published:2015-01-22
  • Supported by:

    The National Natural Science Foundation of China (No. 41171357) Doctoral Scientific Research Fundation of Shandong University of Technology(Nos. 4041-413050 4041-413042) The Fundamental Research Funds for the Central Universities(No. 2012213020212)

摘要:

在常用的多特征提取方法的基础上提出一种附加约束条件的零件轮廓线的多特征提取方法,即在已有方法提取的多特征结果上附加轮廓上圆弧与直线相切、相邻圆弧与圆弧相切等约束条件,对轮廓上的特征进行迭代精确提取,将组成轮廓的各特征如直线、圆弧和圆,进行精确分段识别,获得各特征的参数.最后使用模拟图像和实际的工业零件图像分别进行了试验验证,结果证明约束关系的引入能够有效地提高轮廓线多特征提取的精确度.

关键词: 多特征提取, 视觉检测, 零件轮廓线, 约束条件, 角点检测

Abstract:

Multi-feature extraction of contour is the key process in parts visual measurement. It is presented that a multi-feature extraction method of contours with additional constraints based on existing method to identify the composition primitives of the contour accurately, such as lines, arcs, circles, and acquire the parameter of each primitive. At last, the simulated and the actual industrial parts experimental results demonstrate that the introduction of constraint relations can effectively improve the accuracy of multi-feature extraction.

Key words: multi-feature extraction, visual inspection, parts contour, constraints, corner detection

中图分类号: