›› 2016, Vol. 29 ›› Issue (4): 99-.

• 论文 • 上一篇    下一篇

基于机器视觉的工件定位和识别

陈泽宁,张学习,彭泽荣,钟凯宇   

  1. (广东工业大学 自动化学院,广东 广州 510006)
  • 出版日期:2016-04-15 发布日期:2016-04-26
  • 作者简介:陈泽宁(1989—),男,硕士研究生。研究方向:图像处理与机器视觉。张学习(1978—),男,博士,副教授。研究方向:图像处理与机器视觉。
  • 基金资助:

    广东省工业高新技术领域科技计划基金资助项目(2013B010401028)

Workpiece Location and Recognition Based on Machine Vision

CHEN Zening,ZHANG Xuexi,PENG Zerong,ZHONG Kaiyu   

  1. (School of Automation,Guangdong University of Technology,Guangzhou 510006,China)
  • Online:2016-04-15 Published:2016-04-26

摘要:

根据机器视觉系统具有非接触性、实时性、稳定性及高精度的特点。文中采用机器视觉技术对工件的定位和识别进行了研究。通过求取工件轮廓中心矩和轮廓轴线二阶导数的最小值,求得工件的中心坐标和旋转角度;通过提取模板工件的Hu不变矩与工件轮廓的Hu不变矩,利用轮廓的相似度量函数识别工件类型。实验结果表明,通过提取工件轮廓矩能在旋转的情况下精确地定位和识别工件。

关键词: 工件定位, 工件识别, 轮廓提取, 旋转

Abstract:

The noncontact realtime machine vision system enjoys stability and high precision.In this paper,the machine vision is used to study the localization and recognition of the workpiece.The center coordinates and rotation angles are obtained by solving the contour center moment and the minimum value of the two order derivative of the axis of the workpiece.The contour similarity metric function is used to identify the types of artifacts by extracting the template of artifacts Hu invariant moments and Hu invariant moments of the workpiece contour.The experimental results show that the workpiece can be accurately located and identified by the extraction of the workpiece contour.

Key words: workpiece position;workpiece recognition;contour extraction;rotation

中图分类号: 

  • TP242.6+2