基于机器视觉的智能定位与识别系统及其在激光焊接中的应用研究
Research on Intelligent Positioning and Recognition System Based on Machine Vision and Its Application in Laser Welding
摘要: 在传统生产方式中,机箱混料及螺柱缺失主要依赖人工目检,存在准确性低、生产成本高的问题,会导致错误产品流入焊接工序引发批量报废;同时,机器人焊接依赖预设轨迹,但机箱来料位姿偏差普遍存在,远超激光焊接的精度容限,造成焊偏等工艺缺陷。本文设计了一套基于机器视觉的智能定位与识别系统,并将其应用在机箱的激光焊接领域。工位一通过字符识别、卡尺工具和灰度值标准差算法,实现了对产品线上的机箱混料识别以及目标螺柱有无识别。工位二通过形状模板匹配算法,实现了视觉引导激光焊接。经过大量的实验测试,混料识别和螺柱有无识别的准确性均达到99.5%以上,视觉引导激光焊接的定位误差在±0.1 mm左右,满足现场实际的生产需求。该系统操作简单、准确性高、稳定性强,极大地提高了现场的生产效率,为激光焊接领域的高质量发展提供了强大的技术支撑。
Abstract: In traditional production methods, the mixing of chassis materials and the absence of studs mainly rely on manual visual inspection, which has problems of low accuracy and high production costs. This can lead to incorrect products flowing into the welding process and causing batch scrapping. Meanwhile, robot welding relies on preset trajectories, but the orientation deviation of incoming materials from the chassis is widespread, far exceeding the precision tolerance of laser welding, resulting in process defects such as welding deviation. This paper designs a set of intelligent positioning and recognition systems based on machine vision and applies it to the field of laser welding of chassis. At Workstation One, through character recognition, caliper tools and the standard deviation algorithm of gray values, the identification of mixed materials in the chassis of the product line and the identification of whether there are target studs or not have been achieved. Station Two has achieved vision-guided laser welding through a shape template matching algorithm. After extensive experimental tests, the accuracy of both mixed material identification and stud presence identification has reached over 99.5%, and the positioning error of vision-guided laser welding is approximately ±0.1 mm, meeting the actual production requirements on site. This system is easy to operate and highly accurate. It has strong stability, which greatly improves the on-site production efficiency and provides strong technical support for the high-quality development of the laser welding field.
文章引用:文正彪, 侯士旺, 刘鹏, 王益鑫, 陈相柏. 基于机器视觉的智能定位与识别系统及其在激光焊接中的应用研究[J]. 传感器技术与应用, 2026, 14(2): 278-292. https://doi.org/10.12677/jsta.2026.142028

参考文献

[1] 李君兰. 面向芯片封装的机器视觉精密定位系统的研究[D]: [博士学位论文]. 天津: 天津大学, 2010.
[2] 廖万辉, 李琳. 基于机器视觉的工业机器人定位系统[J]. 微计算机信息, 2009, 25(8): 242-244.
[3] 曾鹏. 基于单目视觉的工件定位与机器人抓取技术研究[D]: [硕士学位论文]. 大连: 大连理工大学, 2017.
[4] 项学智, 马竹山, 周宪坤, 等. 形状匹配的铁路油罐车罐口视觉定位方法[J]. 实验室研究与探索, 2024, 43(5): 5-9.
[5] 张家宁, 刘星龙, 支勇超, 等. 基于机器视觉的工件识别和定位系统设计[J]. 科学技术创新, 2024(4): 64-67.
[6] 陈甦欣, 罗乐文, 赵安宁. 基于机器视觉的电芯绝缘介质定位算法[J]. 合肥工业大学学报(自然科学版), 2024, 47(4): 452-457.
[7] 陈恳. 基于机器视觉的物料分拣工业机器人关键技术研究[D]: [硕士学位论文]. 深圳: 深圳大学, 2017.
[8] Nilsen, M. and Sikström, F. (2025) Integrated Vision-Based Seam Tracking System for Robotic Laser Welding of Curved Closed Square Butt Joints. The International Journal of Advanced Manufacturing Technology, 137, 3387-3399. [Google Scholar] [CrossRef
[9] Eren, B., Demir, M.H. and Mistikoglu, S. (2023) Recent Developments in Computer Vision and Artificial Intelligence Aided Intelligent Robotic Welding Applications. The International Journal of Advanced Manufacturing Technology, 126, 4763-4809. [Google Scholar] [CrossRef
[10] 施航杰. 基于机器视觉引导的电容高精度装配系统研究[J]. 组合机床与自动化加工技术, 2024(5): 71-76+81.
[11] 徐昌军, 曹雏清. 面向水泵叶片自动化上料的视觉定位系统设计[J]. 制造业自动化, 2024, 46(9): 209-214.
[12] 孙克帅, 肖楠, 朱将勇. 基于PLC与机器视觉的自动上料系统[J]. 智慧轨道交通, 2024, 61(6): 93-99.
[13] Manoj, P., Sahoo, A., Maurya, G.S., et al. (2014) Handwritten Character Recognition for English and Telugu Scripts Using Multi-Layer Perceptions (MLP). International Journal of Scientific Engineering and Technology, 3, 730-733.
[14] Kumar, P., Sharma, N. and Rana, A. (2012) Handwritten Character Recognition Using Different Kernel Based SVM Classifier and MLP Neural Network (a Comparison). International Journal of Computer Applications, 53, 25-31. [Google Scholar] [CrossRef
[15] Xiong, J., Wang, D., Yin, J. and Wu, R. (2024) Precise Z-Block Positioning and Dimension Measurement Using Improved Canny Edge Detection and Sub-Pixel Contour Fitting. The Journal of Supercomputing, 81, Article No. 230. [Google Scholar] [CrossRef
[16] Wang, W., Li, L. and Han, Y. (2021) Crack Detection in Shadowed Images on Gray Level Deviations in a Moving Window and Distance Deviations between Connected Components. Construction and Building Materials, 271, Article ID: 121885. [Google Scholar] [CrossRef
[17] 夏群峰, 彭勇刚. 基于视觉的机器人抓取系统应用研究综述[J]. 机电工程, 2014, 31(6): 697-701+710.
[18] 万国扬, 王国峰, 李福东, 等. 基于视觉定位与轨迹规划的机器人打磨工作站[J]. 计算机集成制造系统, 2021, 27(1): 118-127.
[19] 冯立志, 陈翊文, 陈广枢, 等. 基于机器视觉的FPC定位系统设计与实现[J]. 机电工程技术, 2024, 53(10): 163-166+171.
[20] Chen, T., Chang, W., Shen, P. and Tarng, Y. (2009) Examining the Profile Accuracy of Grinding Wheels Used for Microdrill Fluting by an Image-Based Contour Matching Method. Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, 224, 899-911. [Google Scholar] [CrossRef
[21] Wang, R., Yang, K.M. and Zhu, Y. (2023) A High-Precision Mark Positioning Algorithm Based on Sub-Pixel Shape Template Matching in Wafer Bonding Alignment. Precision Engineering, 80, 104-114.
[22] 杜小东, 丁忆凡, 田威, 等. 共形天线表面微小焊盘视觉定位算法设计[J]. 电子机械工程, 2024, 40(2): 53-58+64.
[23] 李欣, 齐家敏, 程昊, 等. 基于机器视觉和PLC的谷糙分级控制系统设计[J]. 现代电子技术, 2024, 47(9): 124-130.
[24] AL-Karkhi, N.K., Abbood, W.T., Khalid, E.A., Jameel Al-Tamimi, A.N., Kudhair, A.A. and Abdullah, O.I. (2022) Intelligent Robotic Welding Based on a Computer Vision Technology Approach. Computers, 11, Article No. 155. [Google Scholar] [CrossRef