基于视觉引导的机器人系统标定及跟随控制的问题研究

发布时间：2018-01-12 20:23

本文关键词：基于视觉引导的机器人系统标定及跟随控制的问题研究　出处：《湖北工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：工业应用中的机器人如果要完全实现智能化,机器视觉引导技术在应用中具有关键的作用。标定环节在视觉引导中起着十分的关键,标定的精度是否达标,直接影响到后期机器人对工件的定位。同时,机器人视觉中标定部分的技术能否达到要求,也是离线编程能否运用到实际中去的重要前提。本文主要研究了工业中各种类型的视觉产品,根据使用中的情况分析其原理和标定方法,其在线跟随功能是其中的一大重点,求解各个坐标系之间关系的变换。研究的主要内容如下:1.视觉引导系统在工业中会运用在各种场合,本文主要研究分析不同场合的系统的工作原理,根据分析结果建立模型,根据所选取的特征点求解工件位置的偏移量。实验表明,同一物体运动中高度有所变化,2.5D视觉系统中获取不同高度时的图像信息,根据图像尺寸的不同计算出高度的变化量。2.建立基于2点的2D视觉系统手眼标定模型。基于2点,顾名思义就需要在工件上选取2个特征点,根据特征点位置的改变,最终得到图像坐标系和工件坐标系之间的变换关系中,求解3个变量。3.建立基于视觉引导的机器人在线跟随技术,建立一个基于单目视觉的工业机器人工件自动识别和智能抓取系统,其中ABB—IRB120型工业机器人作为操作手臂,用CCD相机、运动控制器和工控机搭建了机器人视觉控制平台,通过建立抓取系统的参数化模型给出图像坐标到机器人坐标的转换算法,利用软件中提供的图像处理、模式匹配等方法,利用了卡尔曼滤波算法,在C++环境下进行开发,实现了目标定位和机器人控制两大基本功能,最终控制机器人完成目标动作。
[Abstract]:Machine vision guidance technology plays a key role in the application of industrial robot if it is to be fully intelligent. Calibration plays a very important role in visual guidance and the calibration accuracy is up to standard. At the same time, the calibration technology of robot vision can meet the requirements. It is also an important premise for offline programming to be used in practice. This paper mainly studies various types of visual products in industry and analyzes their principles and calibration methods according to the situation in use. The function of on-line following is one of the key points. The main contents of the research are as follows: 1. Visual guidance system will be used in various situations in industry. This paper mainly studies the working principle of the system in different situations, establishes the model according to the analysis results, and calculates the offset of the workpiece position according to the selected feature points. The image information of different heights is obtained in the 2.5D visual system of the same object. According to the difference of image size, the change of height is calculated. 2. A 2D visual system hand-eye calibration model based on 2 points is established. Based on 2 points, as the name implies, two feature points are selected on the workpiece. Finally, according to the change of feature point position, the transformation relationship between image coordinate system and workpiece coordinate system is obtained, and three variables. 3. The robot online following technology based on vision guidance is established. An industrial robot workpiece automatic recognition and intelligent capture system based on monocular vision is established, in which the ABB-IRB120 industrial robot is used as the operating arm and the CCD camera is used. The robot vision control platform is built by motion controller and industrial control computer. The image coordinate to robot coordinate conversion algorithm is given by establishing the parameterized model of the grab system, and the image processing provided in the software is used. The method of pattern matching is developed in C environment by using Kalman filter algorithm. The two basic functions of target location and robot control are realized. Finally, the robot is controlled to complete the target action.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TP242

【参考文献】