基于机器视觉的自动喷涂系统研究

发布时间：2018-06-02 14:25

本文选题：机器视觉 + 图像处理　；参考：《中国矿业大学》2017年硕士论文

【摘要】：传统喷涂系统存在喷涂效率低、喷涂环境恶劣、喷涂质量无法保障等问题,而由喷涂机器人实现的自动喷涂系统不仅解决了传统喷涂系统存在的问题,而且使用喷涂机器人代替喷涂工人在恶劣环境中进行工作。然而,目前的自动喷涂系统只适用于一些生产量大、产品单一的大型企业,而对于一些生产量小、产品多样的中小企业,自动喷涂系统不能得到很好地应用。基于以上问题,本课题以广州宁基智能系统有限公司的门板喷涂项目为背景设计了一套基于机器视觉的自动喷涂系统。该系统不仅实现了对不同工件的自动检测与喷涂,而且大大降低了生产成本。本文根据系统功能将其分成三大模块:运输模块、机器视觉模块、喷涂机器人模块。运输模块采用两条输送链、提升机以及光电传感器等设备,实现了快速、准确地将工件运输到图像检测区域与喷涂区域的功能。针对喷涂环境的特殊性完成了机器视觉模块硬件设备的选型,并详细的分析了用于工件类型判断与尺寸测量的图像处理算法,如直方图均衡化、去噪、阈值分割、边缘检测以及直线检测等,最后使用传统的标定方法对机器视觉系统进行标定。在喷涂机器人模块中建立了圆锥形喷涂模型以及利用β分布模型建立了涂层累积速率模型,使用涂层的厚度方差作为涂层均匀性的评价函数,根据工件的物理特征提出了一种基于平面梯度的路径规划算法。最后,针对机器视觉模块设计了一套基于HALCON与WinForm的机器视觉检测软件;针对喷涂机器人模块在RobotStudio应用软件中使用IRB5500喷涂机器人实现了工件喷涂路径的仿真与验证。
[Abstract]:The traditional spraying system has some problems, such as low spraying efficiency, bad spraying environment, and the quality of spraying can not be guaranteed. But the automatic spraying system realized by spraying robot not only solves the problems existing in the traditional spraying system. And spray robots are used instead of spray workers to work in harsh conditions. However, the current automatic spraying system is only suitable for some large enterprises with large production and single product. However, for some small and medium-sized enterprises with small production and diversified products, the automatic spraying system can not be applied very well. Based on the above problems, an automatic spraying system based on machine vision is designed based on the door spraying project of Guangzhou Ningji Intelligent system Co., Ltd. The system not only realizes the automatic detection and spraying of different workpieces, but also reduces the production cost. According to the system function, this paper divides it into three modules: transport module, machine vision module and spray robot module. The transport module uses two conveying chains, hoists and photoelectric sensors to transport the workpiece quickly and accurately to the image detection area and the spraying area. According to the particularity of the spraying environment, the hardware equipment selection of machine vision module is completed, and the image processing algorithms used to judge the type of workpiece and measure the dimension are analyzed in detail, such as histogram equalization, denoising, threshold segmentation, etc. Finally, the traditional calibration method is used to calibrate the machine vision system. In the spray robot module, the conical spray model and the coating accumulation rate model are established by using the 尾 distribution model. The thickness variance of the coating is used as the evaluation function of the coating uniformity. According to the physical characteristics of the workpiece, a path planning algorithm based on plane gradient is proposed. Finally, a set of machine vision detection software based on HALCON and WinForm is designed for the machine vision module, and the simulation and verification of the spraying path of the workpiece is realized by using the IRB5500 spraying robot in the RobotStudio application software for the spray robot module.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242;TP391.41

【参考文献】