合肥工业大学校徽 合肥工业大学学报自科版

导航菜单

基于机器视觉的扁线电机定子焊接定位

Stator welding positioning of flat wire motor based on machine vision

期刊信息

合肥工业大学(自然科学版),2025年6月,第48卷第6期:731-736

DOI: 10.3969/j.issn.1003-5060.2025.06.003

作者信息

任永强,韩冲,汪超

(合肥工业大学机械工程学院,安徽合肥230009)

摘要和关键词

摘要: 针对扁线电机生产过程中的定子焊接定位目标多、速度慢等问题,文章提出一种具有旋转不变性的快速多目标模板匹配方法。首先对图像进行预处理,减少干扰;然后采用由粗到精的加速匹配策略。粗匹配阶段,通过图像金字塔算法降低计算成本,利用优化的快速圆投影特征对模板和搜索图像进行匹配,再使用改进的均值漂移算法对匹配点进行底层类聚,完成对多目标的自适应划分,并将类聚中心作为下一阶段的候选点;精匹配阶段,在候选点及其邻域上利用Hu矩进行模板匹配,实现目标的精确定位。实验结果表明,文章所提方法能解决目标旋转问题,可以快速准确地匹配多个目标,且定位精度在3个像素以内,满足实际焊接定位需求。

关键词: 圆投影;均值漂移;Hu矩;模板匹配;多目标

Authors

REN Yongqiang, HAN Chong, WANG Chao

(School of Mechanical Engineering, Hefei University of Technology, Hefei 230009, China)

Abstract and Keywords

Abstract: Aiming at the problems of multiple targets and slow speed of stator welding positioning in the production process of flat wire motors, this paper proposes a fast multi-objective template matching method with rotation invariance. Firstly, the image is preprocessed to reduce the interference, and then an accelerated matching strategy from rough to fine is adopted. In the rough matching stage, the image pyramid algorithm is used to reduce the calculation cost, and the optimized fast ring projection features are used for matching the template and the search image. Then, the improved Mean Shift algorithm is used to perform underlying clustering of the matched points, and the adaptive division of multiple targets is completed. The cluster centers are used as candidate points for the next stage. In the fine matching stage, Hu moments are used to perform template matching on the candidate points and their neighborhoods to achieve precise positioning of the target. Experimental results show that this method can solve the problem of target rotation and match multiple targets quickly and accurately, and the positioning accuracy is within three pixels, which meets the practical requirements of welding positioning.

Keywords: ring projection; Mean Shift; Hu moments; template matching; multi-objective

基金信息

安徽省科技重大专项资助项目(2021d05050002)

个人中心