MATLAB牛奶瓶瓶口破损及残留物自动检测研究

摘要瓶装牛奶生产厂商大都对玻璃牛奶瓶进行回收再利用。回收的牛奶瓶必须在灌装前进行清洗和检测，剔除会影响牛奶质量的次瓶。人工检测速度慢、效率低、漏检率高。而采用自动检测技术能够实现速度快、精度高的自动化检测。为此，本论文对牛奶瓶自动检测技术进行了探讨。7826
本论文首先简要介绍了牛奶瓶检测的背景及现状。在介绍了检测系统的工作流程和总体结构后，对照明系统，成像系统，数字图像处理系统进行了分析和研究，并着重讨论了后期基于MATLAB的图像处理过程。本文采用模版匹配法，通过仿真和分析，选定top-hat和bottom-hat滤波、二值化、图像相减、图像判断的流程确定瓶口、瓶底和瓶壁的合格度。文章最后还介绍基于卷积的图像配准方式和关于瓶底残液残碱的检测方法。
关键词自动检测 MATLAB 图像处理模版匹配
毕业设计说明书（论文）外文摘要
Title   The Research On The Automatic Detection Technology      Of The Milk Bottle’s Breakage And Residue
Abstract
Most of the manufacturers of bottled milk have the bottles recycled. Before filling, the glass bottles must be cleaned and tested to get rid of the defective items which will affect the quality of the milk. Manual inspection is of low velocity, poor efficiency and high false detecting rate. While the fast and high-precision automated inspection can be achieved by the automatic detection technology. To this end, the milk bottle automatic detection techniques are discussed.
This paper briefly describes the background and current situation of the milk bottle detection. After the introduction of the workflow and the overall structure of the detection system, we analyze and research the lighting systems, the imaging systems, and the digital image processing system, and focus on the image processing based on MATLAB. The template-matching method is the key point of this paper. By simulation and analysis, top-hat/bottom-hat filter, binaryzation, image subtraction and image distinguish are selected as the workflow to detect the quality of the bottleneck, the bottom and the sidewall. At the terminal of the paper, the convolution-based image registration and the detection of the residual alkali and liquid on the bottom are discussed.
Keywords automatic-detection MATLAB image-processing template-matching
目录

1 引言（或绪论）   1
1.1 课题研究背景   1
1.2 目前现状   1
1.3 MATLAB图像处理工具箱简介   2
1.4 本文的研究内容及章节安排   2
2 系统设计   3
2.1 总体框图   3
2.2 照明系统：   3
2.3 成像系统：   7
2.4 数字图像处理：   8
2.4.1 数字图像处理的主要研究内容：   8
2.4.2 数字图像处理的实现算法：   9
3 瓶口完整性判别   12
3.1 增强模版瓶口图像对比度   12
3.2 模版瓶口图像二值化   13
3.3 破损瓶口图像处理   15
3.4 处理后图像相减   16
3.5 再次二值化   17
4 瓶底和瓶壁污物的判别   18
4.1 模版瓶底图像的TOP-HAT和BOTTOM-HAT滤波   18 MATLAB牛奶瓶瓶口破损及残留物自动检测研究:http://www.youerw.com/tongxin/lunwen_5922.html