基于图像分析的打印缺陷在线检测方法研究与实现
毕业论文关键词 打印质量 图像分析 在线检测 机器视觉
Title Research and Implementation on Online Detection Method of Printing Defect Based on Image Analysis
Abstract
Laser printing has become an indispensable part in people's daily life，and the
quality of printing also calls attention. However, due to the subjective factors,
the problems of traditional manual testing method are as follows: high labor
intensity and labor cost, uncontrolled artificial error. Therefore, with the
development of the software and hardware of computer, it is practical to design
and develop an online detection method.
In this paper, research and implementation on online detection method of printing
defect has been performed by the way of image analysis. First, images were
real-timely captured by machine vision to form a local gallery with a certain
number. Second, design different detection method according to different problems
of the products: (1) project the text to the edge, in order to judge whether it
is inclined; (2) measure whether it is out of ink by the average of the column
pixels (without the background points). Finally, reach the detection accuracy of
95% after being tested by the local gallery.
Keywords Printing Quality, Image Analysis, Online Detection, Machine Vision
目 次
1 引言 .. 1
1．1 课题研究的背景和意义 . 1
1．2 打印缺陷检测的研究现状 .. 1
1．3 本文的主要内容和章节安排 2
2 数字图像处理 3
2. 1 数字图像的定义 . 3
2. 2 本文使用的处理方法 3
2.2.1 图像平滑 .. 3
2.2.2 边缘检测 .. 4
2.2.3 图像变换 .. 8
2．3 数字图像处理的优点 .. 10
2．4 本章小结 11
3 硬件系统设计 .. 12
3. 1 机器视觉检测概述 . 12
3. 2 激光打印原理 .. 12
3. 3 硬件平台介绍 .. 14
3. 4 本章小结 15
4 打印缺陷检测 .. 16
4．1 倾斜检测 . 16
4．2 缺墨检测 . 17
4．3 人机界面 . 19
4．4 本章小结 . 20
5 实验 . 21
5．1 检测结果 . 21
5．2 优化过程 . 29
5．3 本章小结 . 32
结 论 .. 33
致 谢 .. 34
参 考 文 献 35
1 引言
1．1 课题研究的背景和意义
随着现代科技和计算机技术的发展，各种打印品和印刷品在人们的日常生活中已经占据
上一篇：基于深度学习的时序模式识别