【作者】 董辉；

【导师】 王洪瑞；

【作者基本信息】 河北大学 ， 控制理论与控制工程， 2010， 硕士

【摘要】 随着工业自动化的发展,生产与加工的自动化程度越来越高,机器人视觉技术在自动化生产中也变得越来越重要。双目立体视觉技术是机器人视觉领域的一个重要课题,是智能机器人的重要标志。在查阅大量资料的基础上,本文重点对双目立体视觉的基本理论、方法和相关技术进行了系统研究;并在此基础上建立了一套视觉伺服控制机器人系统,在此系统上可以开展机器人视觉伺服控制理论及技术的研究,具体工作包括以下几个方面:1、根据本文的研究目的和实际具有的设备环境,应用Bumblebee2双目相机、高精度X-Y平台、工控机、松下的MSDA043A1A交流伺服电机驱动器等搭建了双目立体视觉伺服系统。2、利用中值滤波和阈值分割的方法提取出目标特征,为三维重建提供了图像特征点的像素坐标数据。3、三维重建是双目立体视觉的最终目的,本文主要利用左右两幅图像的匹配点以及所求得的投影矩阵,对目标点进行三维空间坐标重建。4、利用面向对象设计与实现技术,选择C++语言与VC++开发环境,开发了图像采集、图像目标提取、三维重建以及机器人控制软件,实现对操作平台上的运动目标进行跟踪、捕捉或接取等任务。本文在Visual C++集成开发环境中基于OpenCV编程实现了系统的功能。完成了实时处理系统的算法设计。试验结果说明本系统能够实时捕捉到单个运动目标的三维运动信息,可为后续功能的开发打下坚实的基础。更多还原

【Abstract】 With the development of the industry automation, and the improvement of the automation level of production, the robot vision becomes more and more important in the automatic production. Binocular stereo vision technology is an important subject in the field of robot visions, and it is also the important symbol of the intelligent robots.By referring to a great deal of the information available, binocular stereo vision’basic theory, method and related techniques have been discussed in this paper, and based on which, a robot visual serving control system is constructed, in this system the visual servo control theory and technology research can be carried out. The main work in this paper is summarized as follows:(1) According to the aim and the device possessed, using the Bumblebee2 binocular camera, X-Y platform, IPC and the MSDA043A1A Panasonic AC servo motor drive to construct the binocular stereo vision system.(2) Using median filter and threshold to extract the target feature, and provide the pixel coordinates for stereo matching.(3) The final purpose of binocular stereo vision is 3D reconstruction. The target point can be reconstructed by using the matching points of left and right image and projection matrix.(4) Adopting object-oriented design and realized technology; selecting C++ language and Visual C++ development environment, this paper develops related software including image capture, target extract, 3D reconstruction and robot control system, the mission that track, capture or access the moving targets is implemented on the platform.Visual C++ environment and OpenCV tools are used. The design of algorithms and hardware has been achieved in this paper. The experiments show that the system can capture real-time three-dimensional motion information of single moving object. It is the foundation of other function of the robot.更多还原