本文關鍵詞：基于視覺引導的機器人系統(tǒng)標定及跟隨控制的問題研究　出處：《湖北工業(yè)大學》2017年碩士論文　論文類型：學位論文

  更多相關文章： 工業(yè)機器人 視覺引導 手眼標定 在線跟隨 卡爾曼濾波

【摘要】：工業(yè)應用中的機器人如果要完全實現(xiàn)智能化,機器視覺引導技術在應用中具有關鍵的作用。標定環(huán)節(jié)在視覺引導中起著十分的關鍵,標定的精度是否達標,直接影響到后期機器人對工件的定位。同時,機器人視覺中標定部分的技術能否達到要求,也是離線編程能否運用到實際中去的重要前提。本文主要研究了工業(yè)中各種類型的視覺產(chǎn)品,根據(jù)使用中的情況分析其原理和標定方法,其在線跟隨功能是其中的一大重點,求解各個坐標系之間關系的變換。研究的主要內(nèi)容如下:1.視覺引導系統(tǒng)在工業(yè)中會運用在各種場合,本文主要研究分析不同場合的系統(tǒng)的工作原理,根據(jù)分析結果建立模型,根據(jù)所選取的特征點求解工件位置的偏移量。實驗表明,同一物體運動中高度有所變化,2.5D視覺系統(tǒng)中獲取不同高度時的圖像信息,根據(jù)圖像尺寸的不同計算出高度的變化量。2.建立基于2點的2D視覺系統(tǒng)手眼標定模型�；�2點,顧名思義就需要在工件上選取2個特征點,根據(jù)特征點位置的改變,最終得到圖像坐標系和工件坐標系之間的變換關系中,求解3個變量。3.建立基于視覺引導的機器人在線跟隨技術,建立一個基于單目視覺的工業(yè)機器人工件自動識別和智能抓取系統(tǒng),其中ABB—IRB120型工業(yè)機器人作為操作手臂,用CCD相機、運動控制器和工控機搭建了機器人視覺控制平臺,通過建立抓取系統(tǒng)的參數(shù)化模型給出圖像坐標到機器人坐標的轉換算法,利用軟件中提供的圖像處理、模式匹配等方法,利用了卡爾曼濾波算法,在C++環(huán)境下進行開發(fā),實現(xiàn)了目標定位和機器人控制兩大基本功能,最終控制機器人完成目標動作。
[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.
【學位授予單位】：湖北工業(yè)大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TP391.41;TP242

【參考文獻】

相關期刊論文 前10條

1 周蓉;劉波;;基于目標區(qū)域預檢測模型的粒子濾波小目標跟蹤算法[J];計算機應用;2013年S2期

2 王才東;王新杰;王輝;陳鹿民;;六自由度機器人控制系統(tǒng)設計與研究[J];計算機仿真;2013年09期

3 吳文祥;朱世強;王宣銀;劉華山;;基于摩擦模糊建模與補償?shù)臋C器人低速控制[J];電機與控制學報;2013年08期

4 蔡自興;郭t，

本文編號：1415850