一种基于Mark点的点胶机器人视觉目标定位方法

PDF下载HTML

Title:

A method of vision target localization for dispensing robot based on mark point

作者:

彭刚^1,2, 熊超^1,2, 夏成林², 林斌³

1. 华中科技大学 自动化学院, 湖北 武汉 430074;
2. 华中科技大学 图像信息处理与智能控制教育部重点实验室, 湖北 武汉 430074;
3. 深圳市海思科自动化技术有限公司, 广东 深圳 518109

Author(s):

PENG Gang^1,2, XIONG Chao^1,2, XIA Chenglin², LIN Bin³

1. School of Automation, Huazhong University of Science and Technology, Wuhan 430074, China;
2. Key Laboratory of Image Processing and Intelligent Control of Education Ministry, Huazhong University of Science and Technology, Wuhan 430074, China;
3.

关键词:

PCB; 点胶机器人; Mark点; 辅助; 视觉定位; 几何特征; 改进型模板匹配; 不变性

Keywords:

PCB; dispensing robot; Mark point; assistant; visual localization; geometric features; improved template matching; invariance

分类号:

TP242

DOI:

10.11992/tis.201705010

摘要:

针对工业生产中PCB点胶机器人的视觉定位问题，提出了一种基于Mark点辅助的视觉定位算法。分析了传统模板匹配、Sift、Surf等算法在Mark点识别与定位中的不足，同时考虑到Mark点所具有的规则几何特征以及算法对于实时性的要求，提出了一种基于Mark点几何特征的改进型模板匹配算法。实验结果表明，这种基于Mark点几何特征的改进型模板匹配算法具有良好的平移、缩放、旋转不变性，能够准确识别并定位Mark点，从而实现对PCB上相关点胶目标点的定位，并满足工程可靠性和实时性的要求。

Abstract:

To address the problem of visual localization of polychlorinated biphenyl (PCB)-dispensing robots in industrial production, a visual localization algorithm based on mark point is proposed in this paper. The results of traditional template matching, Sift and Surf algorithms, were compared and it was found that these algorithms are inadequate for the identification and location of Mark point. Meanwhile, the geometry features of the Mark point and the real-time requirements of the algorithm were considered so that an improved template-matching algorithm based on Mark point geometry features was realized. Experimental results show that this improved template-matching algorithm based on Mark point geometry features can keep good invariance under translation, scale, and rotation, which can accurately identify and locate the mark points, and then indirectly locate the PCB target points while satisfying engineering reliability and real-time constraints.