汽车挡风玻璃定位与支架粘合视觉检测系统的设计与实现

发布时间：2018-03-28 05:10

本文选题：机器视觉　切入点：挡风玻璃定位　出处：《电子测量与仪器学报》2017年03期

【摘要】：针对企业提出的汽车前挡风玻璃自动制造需求,设计了一套基于机器视觉的前挡风玻璃定位与支架粘合检测系统。阐述了系统的整体设计方案,详细介绍了电气控制部分、机器人控制部分和机器视觉部分。系统采用底部背光与双侧面条形光相结合的光学控制策略。针对产品的特征,提出了自适应阈值图像分割算法和基于改进的随机霍夫变换特征提取算法,实现汽车挡风玻璃定位和支架粘合质量检测。实验结果表明,设计的系统能够很好地降低粘合支架的次品率,次品率仅为4%,满足了企业生产现场的精度和时间要求,实现粘合精度小于0.5 mm,智能制造系统生产效率为人工的4.5倍,大大提高企业的生产效率。
[Abstract]:According to the auto manufacturing requirement of automobile front windshield, a set of front windshield positioning and bracket bonding detection system based on machine vision is designed. The overall design scheme of the system is described, and the electrical control part is introduced in detail. Robot control part and machine vision part. The system adopts optical control strategy which combines bottom backlight and double side strip light. An adaptive threshold image segmentation algorithm and an improved feature extraction algorithm based on stochastic Hough transform are proposed to locate the windshield and detect the adhesion quality of the frame. The experimental results show that, The system designed can reduce the defective rate of the adhesive bracket very well. The defective rate is only 4%, which meets the precision and time requirement of the production site of the enterprise. The adhesive precision is less than 0.5 mm, and the production efficiency of the intelligent manufacturing system is 4.5 times that of the artificial one. Greatly improve the production efficiency of enterprises.
【作者单位】：湖南大学电气与信息工程学院;机器人视觉感知与控制技术国家工程实验室;湖南科技大学信息与电气工程学院;
【基金】：国家自然科学基金(61072121,61271382) 国家科技支撑计划(2015BAF13B00)资助项目
【分类号】：U468.2;TP391.41

【相似文献】