CASIA OpenIR  > 毕业生  > 硕士学位论文
Alternative TitleDetection of Two-Dimension Bar Code in Complex Scene
Thesis Advisor王欣刚
Degree Grantor中国科学院研究生院
Place of Conferral中国科学院自动化研究所
Degree Discipline计算机应用技术
Keyword二维条形码 检测 纹理特征 机器学习 2d Bar Code Detection Texture Feature Machine Learning
Abstract由于二维条形码信息存储量大、容错性强、制作与携带方便,被广泛用于物流、销售、娱乐等各个行业。目前,二维条形码译码技术正由传统的光电扫描方式向图像译码方式过渡。图像译码技术即二维条形码识别技术是由图像采集设备对二维条形码进行成像,然后利用数字图像处理方法抽取条码图像中的编码信息,并对编码信息进行解码进而得到隐藏在二维条形码中信息的计算机信息处理方法。此方法具有成本低、译码快、操作简便等特点,近年来受到越来越多的关注。但在实际应用中,由于复杂背景复,条码形态变化以及图像处理因素,图像条码识别技术的应用受到了限制。 在上述行业背景下,本文对二维条码图像形态以及各种相应复杂场景进行了分析,提出一种鲁棒性强、适用性广的二维条码识别算法。本文完成的具体工作和取得的研究成果主要包括以下几个方面: (1) 分析了二维条形码的发展,行业标准以及几种常见的复杂背景以及干扰源,总结了基于图像处理的二维条码识别的几种主要思路和已取得的研究成果,最后对本文内容予以简要叙述。 (2) 概述了本研究提出的二维条形码识别算法流程。重点叙述了复杂背景下二维条形码的分割与定位。本研究充分利用了条码区域的边缘密集,角点集中等特性,并结合区域生长的分割策略,实现了一种基于图像处理的二维条形码区域分割方法。实验表明,本文所提出的技术路线可以很好的适应复杂场景下不同形态的二维条码的识别。 (3) 在本文的研究基础上,深入挖掘了二维条码的本质,尝试将纹理特征与机器学习引入到二维条码识别框架中,以改善系统性能。
Other Abstract2D bar code is a useful tool which can store a large amount of information, with strong fault tolerant ability. It is easy to crate and transform these codes, so 2D bar codes are widely used in logistics, marketing and entertainment industry. In present, the decoding technology is transiting from traditional photo-electric scanning method to image-based method. Image-based method reads 2D bar code through camera and abstracts bit stream from the image, then, decodes the data to get the information hidden in the 2D bar code. More and more people are concerning research of this method because of its characteristic of low cost, fast and easy operation. However, in practice, the application of image-based decoding technology is confined for complex background, variance codes modality and other factors of image processing. Concrete work and major research results in this paper are as followed: (1) Carried out a detailed analysis on the development of 2D bar codes, industry standard and several common complex background and interference sources. Summarized previous major technical routes and research results of image-based 2D bar code recognition algorithm. In the end, briefly reviewed the content of this paper. (2) Outlined the flow of 2D bar code recognition algorithm offered in this paper. And then focused on segmentation and location of 2D bar codes in images with complex background. This research made full use of the characteristics of bar code image, such as denseness of edges, concentrating of corner points. It also combined segmentation strategy of region growing to implement a region segmentation method based on image processing. Experiments showed that the technical route offered by the paper is adaptive in recognizing different forms of 2D bar codes in complex scenes. (3) Based on the research of this paper, dug deeply into the nature of 2D bar codes and tried to introduce features of texture and machine learning methodology into 2D bar codes recognition framework for improving performance of the system.
Other Identifier200728014628091
Document Type学位论文
Recommended Citation
GB/T 7714
刘东昌. 复杂背景下二维条形码的识别[D]. 中国科学院自动化研究所. 中国科学院研究生院,2011.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20072801462809(2367KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[刘东昌]'s Articles
Baidu academic
Similar articles in Baidu academic
[刘东昌]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[刘东昌]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.