摘　要
在現(xiàn)代工業(yè)自動化生產(chǎn)中，自動灌裝機(jī)械是其重要的組成部分，然而目前許多企業(yè)的灌裝機(jī)械多采用人工灌裝或半自動灌裝，尤其是偏心桶的灌裝，應(yīng)用更加稀少。對于長期操作灌裝有毒或者是腐蝕性和滲透性較強(qiáng)的液體的工人的身體健康帶來極大的傷害。因此，本課題研究開發(fā)出基于機(jī)器視覺的自動灌裝系統(tǒng)極其有應(yīng)用價(jià)值。
本文簡要的概述了機(jī)器視覺技術(shù)的發(fā)展歷程、虛擬儀器概念、LabVIEW和圖像處理包IMAQ Vision以及基于PC的機(jī)器視覺系統(tǒng)的體系結(jié)構(gòu)，詳細(xì)介紹了圖像的預(yù)處理、圖像識別和定位。并且對圖像識別和定位進(jìn)行了詳細(xì)的分析：包括利用模板匹配對灌裝桶桶口進(jìn)行粗定位；直接在灰度圖像中利用圓檢測對桶口進(jìn)行二次精確定位；利用信號控制步進(jìn)電機(jī)帶動相機(jī)沿著載物臺上下移動固定距離，經(jīng)過成像模型透視投影，再通過幾何運(yùn)算得到灌裝桶的高度，從而得到桶口的三維坐標(biāo)；在伺服運(yùn)動模塊中能夠在步進(jìn)電機(jī)出現(xiàn)失步的情況下通過對加注槍的識別與定位，對系統(tǒng)的進(jìn)行微調(diào)，最終實(shí)現(xiàn)灌裝桶的正常加注，增加了系統(tǒng)的可控性。
本文根據(jù)系統(tǒng)的開發(fā)過程把系統(tǒng)分為總體方案設(shè)計(jì)、算法設(shè)計(jì)以及系統(tǒng)軟件實(shí)現(xiàn)等幾部分。結(jié)合虛擬儀器的特點(diǎn)和軟件設(shè)計(jì)的要求，采用模塊化的思想設(shè)計(jì)軟件，可以把軟件劃分為用戶登陸管理、攝像機(jī)標(biāo)定、圖像采集和處理、圖像識別與定位、注射槍微調(diào)處理、伺服運(yùn)動控制、數(shù)據(jù)庫等模塊，并對每個(gè)模塊一一進(jìn)行了介紹。詳細(xì)的介紹了LabVIEW與數(shù)據(jù)庫連接方法、LabVIEW和西門子S7-200PLC的通訊及PPI協(xié)議等系統(tǒng)實(shí)現(xiàn)的方法。
在自動灌裝系統(tǒng)開發(fā)過程中，利用LabVIEW、Vision assistant進(jìn)行編程，實(shí)現(xiàn)了圖像的采集、分析、識別與定位；對測量結(jié)果進(jìn)行存儲。通過實(shí)驗(yàn)表明，所設(shè)計(jì)系統(tǒng)完全可以實(shí)現(xiàn)具體的灌裝任務(wù)，取得了滿意的結(jié)果。

關(guān)鍵詞　LabVIEW；機(jī)器視覺；三維坐標(biāo)；模板匹配；自動灌裝


Abstract
Automatic filling machinery is an important part of modern industrial automation production. But now most the filling machine enterprises mainly use artificial filling or semi-automatic filling. Especially the eccentric barrels' filling, its application is scarcer than the other. For long-term operation workers on toxic, corrosive or permeability strong liquid filling, it can brings the enormous damage to workers' health. So the subject research and develop of the automatic filling system based on machine vision have important application value.
In this paper, we summarize the development course of machine vision technology, the concept of virtual instrument, and the soft of LabVIEW and the image processing bag IMAQ Vision, and the system structure of machine vision based on PC. And give a detail analysis about the image recognition and positioning: including the rough positioning of filling barrel's mouth using template matching method; The accurate location of filling barrel's mouth using the round detection algorithm; We can use signal control stepping motor to drive the camera moving fixed distance along the objective shelf, through the perspective projection of imaging model, and get the height of the filling bucket through the geometric operation, then we can get 3D coordinate of the mouth of bucket; We can achieve identification and orientation of the filling gun when the stepping motor appear in out-of-step, then fine-tuning the system, finally realize normal charging of filling barrel, and increase the controllability of system.
According to the system's development process, the paper divides the system into the following parts: scheme design, algorithm design and system software realizing, etc. We can combine the characteristics of virtual instrument and the requirements of software design, to design software using the idea of modular. The software can be divided into users log on management, video camera calibration, image acquisition and processing, image recognition and position, the fine-tuning of filling gun, servo motor control, database, and other modules, and each module are introduced in detail. The paper also give a detail about the connection method of database and LabVIEW, the communication of LabVIEW and Siemens S7-200 PLC and PPI agreement and so on.
In this paper,the software development of the automatic filling system adopts labview and Vision assistant,which are powerfull image processing and analysis vision toolkits.The software realizes the function of image collection,image process, image recognition and position, and storage the result of the inspection system.At last, experiments proved that the method discussed can be achieved with satisfactory results.

Key Words　LabVIEW; Machine vision; 3D coordinate; Template matching; Automatic filling


目　錄
摘　要 I
ABSTRACT III
第1章　緒　論 1
1.1　課題研究的背景及意義 1
1.2　國內(nèi)外機(jī)器視覺技術(shù)的研究現(xiàn)狀 1
1.2.1 國外機(jī)器視覺研究現(xiàn)狀 1
1.2.2 國內(nèi)機(jī)器視覺研究現(xiàn)狀 2
1.3 論文的思路及內(nèi)容與論文結(jié)構(gòu) 2
1.3.1 主要研究思路 2
1.3.2 主要研究內(nèi)容 3
1.3.3 論文結(jié)構(gòu) 3
第2章　機(jī)器視覺技術(shù)及虛擬儀器 5
2.1 機(jī)器視覺技術(shù) 5
2.1.1 機(jī)器視覺的概念 5
2.1.2 機(jī)器視覺的應(yīng)用領(lǐng)域 5
2.1.3 機(jī)器視覺系統(tǒng)的硬件組成 6
2.1.4 機(jī)器視覺中的圖像處理技術(shù) 7
2.2 虛擬儀器技術(shù) 8
2.2.1 虛擬儀器 8
2.2.2 LabVIEW語言 12
2.2.3 實(shí)用工具軟件包及IMAQ Vision圖像處理模塊 13
2.3本章小..