摘要该论文结合了现在工业生产的实际情况，而新研发的一个自动物料分拣机器人系统。在自动物料分拣系统中存在机器人控制系统参数确定和分拣对象识别两个突出的技术难点，成为限制工业机器人进行物料自动分拣的技术关键。本文主要解决的是，使机器人系统具有机器视觉识别目标并提取需求特征的问题。
该论文设计了一个自动物料分拣机器人视觉定位系统，其主要工作包括提取物料图像、对图像进行处理以及目标自动识别等内容。应用三文建模软件 Solidworks 完成三轴直角坐标机器人的机械本体设计和建模，并完成驱动系统的计算选型。为了使机器人能够分拣特定形状特征的工件，本文进行了基于机器视觉技术目标自动识别的研究。采用中值滤波对图像预处理、并融合 Canny 算子对图像进行边缘提取和随机 Hough 变换提取所需圆特征实现圆形工件的识别。最后，建立物理样机并调试运行。包括硬件系统的设计选型和利用编程 C# 语言在Visual Studio 2005软件平台上开发软件系统。通过实验演示三轴直角坐标机器人自动抓取与分拣直径约为 50mm 圆形工件的过程。实验结果表明系统能够自动识别指定工件并完成分拣作业。
本文利用虚拟样机技术设计的自动物料分拣机器人系统实现了控制系统参数的自整定和基于机器视觉的圆形工件分拣目标自动识别，可以促进先进的机器人技术在物料自动分拣领域的应用，提高社会生产率。22437
毕业论文关键词：机器人；自动分拣；视觉定位
Abstract
This paper combines the present day realities of the industrial production, and a newly developed automatic materials sorting robot system. Robot control system in automatic material sorting system parameters identification and sorting objects identified two distinct technical difficulties, become key industrial robots for automatic sorting of materials technology. This article is mainly to solve, robot system with machine vision recognition objectives demand characteristics and extraction problems.
The paper designed an automatic material sorting robot Visual location system, its tasks include deliveries of material images, image processing and automatic target recognition, and more. Using Solidworks modeling software to complete three-axis Cartesian robot's mechanical body design and modeling and calculation of the complete drive system selection. To enable the robot to sort specific shape characteristic of artifacts, this article was based on the machine vision technique to automatic target recognition research. Using median filtering on image processing, and integration of image edge detection of Canny operator and randomized Hough transform to extract the desired feature to achieve recognition of round workpieces. Finally, building a physical prototype and commissioning. Hardware system design and selection and use of the c # programming language in Visual Studio 2005 software platform to develop software systems. Through experiments in three-axis linear robots automatically crawl and pickup is about 50mm in diameter round workpieces. Experimental results show that the system is able to automatically recognize the specified parts and complete sorting jobs.
Using virtual prototyping design in automatic material sorting robot system control parameters self-tuning and circular workpieces pickup automatic target recognition based on machine vision can promote advanced robotic technology in the application of automatic sorting material to improve social productivity.
Key words:  robot ； automatic sorting;  Visual positioning
目 录
摘要 . .   1
Abstract  .  2
第一章 绪论   .  4
1.1 研究背景及意义.  .4
       1.2 国内外自动物料分拣机器人的发展状况. 4 自动物料分拣机器人的视觉定位技术研究+CAD图纸:http://www.751com.cn/jixie/lunwen_15067.html