基于機(jī)器視覺(jué)的注塑制品缺陷檢測(cè)系統(tǒng)研究

摘要

隨著計(jì)算機(jī)和機(jī)器視覺(jué)技術(shù)的不斷發(fā)展，基于機(jī)器視覺(jué)的產(chǎn)品檢測(cè)技術(shù)正逐漸成為研究的熱點(diǎn)。基于機(jī)器視覺(jué)的產(chǎn)品檢測(cè)技術(shù)是指以機(jī)器視覺(jué)為手段獲取被測(cè)物體圖像，并將其與己知的標(biāo)準(zhǔn)進(jìn)行比較，從而確定被測(cè)物體的質(zhì)量狀況的過(guò)程，它具有非接觸、速度快、柔性好等突出優(yōu)點(diǎn)，相比于傳統(tǒng)檢測(cè)技術(shù)具有更為廣闊的應(yīng)用前景?；诖?，本文依托東北大學(xué)流程工業(yè)綜合自動(dòng)化重點(diǎn)實(shí)驗(yàn)室基金項(xiàng)目，針對(duì)機(jī)器視覺(jué)技術(shù)在注塑制品缺陷檢測(cè)中的應(yīng)用展開(kāi)研究。

在調(diào)研注塑生產(chǎn)過(guò)程、查閱大量文獻(xiàn)的基礎(chǔ)上，本文完成了基于機(jī)器視覺(jué)的注塑制品質(zhì)量檢測(cè)系統(tǒng)的軟硬件設(shè)計(jì)，并針對(duì)產(chǎn)品圖像采集和處理過(guò)程中遇到的問(wèn)題提出了相應(yīng)的解決方案。

針對(duì)所獲取的圖像存在背景、噪聲等干擾信息，不適于直接進(jìn)行缺陷檢測(cè)的問(wèn)題，本文研究了相關(guān)的圖像處理算法。對(duì)于產(chǎn)品圖像存在背景干擾的情況，提出一種閾值分割與差影相結(jié)合的方法，實(shí)現(xiàn)背景的完全消除；針對(duì)傳統(tǒng)線性濾波以及中值濾波方法中存在的不足，提出一種新的濾波方法，該方法不僅增強(qiáng)了背景分割算法對(duì)于外界環(huán)境變化的適應(yīng)能力，而且提高了算法的實(shí)用性。

在完成圖像處理之后，本文針對(duì)注塑制品常見(jiàn)形狀以及紋理缺陷的特征提取進(jìn)行了研究。一方面根據(jù)系統(tǒng)對(duì)檢測(cè)速度的要求，提出一種快速預(yù)檢測(cè)和缺陷細(xì)節(jié)信息分析識(shí)別相結(jié)合的檢測(cè)思路，在保證缺陷信息完整的情況下，提高檢測(cè)速度；另一方面針對(duì)傳統(tǒng)方法在紋理特征提取中存在的分類效率低下等問(wèn)題，提出一種新的特征組合方法，該方法有效降低了特征向量的維數(shù)，在保證識(shí)別準(zhǔn)確率的情況下，提高分類效率。

最后，本文根據(jù)注塑產(chǎn)品多缺陷種類并存的特點(diǎn)設(shè)計(jì)了基于多分類支持向量機(jī)的特征分類方法。綜合應(yīng)用上述方法，實(shí)現(xiàn)了基于機(jī)器視覺(jué)的注塑制品缺陷檢測(cè)系統(tǒng)的雛形，獲得了較高的檢測(cè)正確率，較好的滿足了注塑制品檢測(cè)的要求。

關(guān)鍵詞：機(jī)器視覺(jué)；注塑制品；圖像處理；特征提取；多分類支持向量機(jī)

 

Research on Injection Product Defects Detection System Based on Machine Vision 

Abstract

The technique of product inspection based on machine vision has been propelled by the development of computer science and machine vision in most recent years, and it seizes more and more researchers’ attention. By snatching the images of the produce and comparing them with standard one, machine vision based inspection can give a real-time eva luation on the quality of the product without contiguity. For these merits, this study focuses on the defects detection of injection molding machine (IMM) product based on machine vision. This research is sponsored by NEU key laboratory of process industry automation fund.

Firstly, by researching the process of injection production and analyzing relative literatures, this thesis finishes the hardware and software design of defects detection system and solves the problems in capturing and processing images. 

The background and noise in the captured image makes the defect detection much harder. A method combining threshold and image subtraction is proposed to segment the background from the object image entirely. To filtering the noise and enhance the robustness of the segmentation method, a new filter algorithm is presented which provides a better result than the traditional methods.

The research on defect detection feature extraction consists of two parts. For detecting shape defects, a rapid-inspecting method is proposed under the condition that the defect’s information is reserved entirely. For detecting texture defects, by regroup the texture feature vectors, a new method is presented to enhance the efficiency of classification.   

At last, a multi-class classification support vector is designed for the multi-class defects detection. The basic structure of IMM products defect detection system based on machine vision is completed by using methods proposed above. The test results show that this system demonstrate a high detecting precision.

Keywords：Machine Vision; Injection Product; Image Processing; Feature Extraction; Multi-class Classification Support Vector Machine

 

目錄

畢業(yè)設(shè)計(jì)（論文）任務(wù)書 I

摘要 II

ABSTRACT III

第一章 緒論 1

1.1 機(jī)器視覺(jué)檢測(cè)技術(shù)概述 1

1.2 機(jī)器視覺(jué)檢測(cè)的研究概況 2

1.3 機(jī)器視覺(jué)檢測(cè)技術(shù)的應(yīng)用 3

1.4 課題背景及本文主要工作 5

第二章 系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn) 7

2.1 系統(tǒng)總體結(jié)構(gòu) 7

2.2 硬件系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn) 7

2.2.1 硬件系統(tǒng)設(shè)計(jì) 8

2.2.2 系統(tǒng)關(guān)鍵設(shè)備選型 8

2.2.3 硬件系統(tǒng)的實(shí)現(xiàn) 10

2.3 軟件系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn) 11

2.3.1 軟件系統(tǒng)基本框架 11

2.3.2 圖像的采集與顯示模塊 12

2.3.3 系統(tǒng)通訊模塊 15

2.3.4 輔助功能模塊 16

2.4 本章小結(jié) 17

第三章 注塑制品圖像處理 19

3.1 注塑制品圖像處理總述 19

3.2 注塑制品圖像背景分割 19

3.2.1 傳統(tǒng)背景分割方法 20

3.2.2 傳統(tǒng)方法在注塑制品背景分割中的應(yīng)用分析 21

3.2.3 注塑制品背景分割方法設(shè)計(jì) 23

3.3 注塑制品圖像濾波 24

3.3.1 傳統(tǒng)濾波方法 25

3.3.2 傳統(tǒng)濾波方法應(yīng)用效果分析 27

3.3.3 注塑制品圖像濾波方法設(shè)計(jì) 27

3.4 本章小結(jié) 29

第四章 結(jié)束語(yǔ) 30

參考文獻(xiàn) 32

致謝 34