基于灰度直方图和谱聚类的文本图像二值化方法

吴锐; 黄剑华; 唐降龙; 刘家锋

doi:10.3724/SP.J.1146.2008.01283

基于灰度直方图和谱聚类的文本图像二值化方法

doi: 10.3724/SP.J.1146.2008.01283

基金项目:

国家自然科学基金(60672090)资助课题

计量
- 文章访问数: 3848
- HTML全文浏览量: 84
- PDF下载量: 1652
- 被引次数: 0
出版历程
- 收稿日期: 2008-10-09
- 修回日期: 2009-03-17
- 刊出日期: 2009-10-19

Method of Text Image Binarization Processing Using Histogram and Spectral Clustering

摘要

摘要: 在自动文本提取中，经定位获得的字符区域需二值化后方能有效识别，由于背景的复杂，常用的阈值化方法不能有效分割自然环境下的字符图像。该文提出了一种基于谱聚类的图像二值化方法，该方法利用规范化切痕(Normalized cut, Ncut)作为谱聚类测度，结合灰度直方图计算相似性矩阵，并通过实验确定最佳的直方图等级数，与通常基于像素级相似矩阵相比，算法的空间复杂度和计算复杂性都大为降低。实验结果表明，针对自然场景下的字符图像，该文方法的二值化结果优于常用的阈值分割结果。
- 图像处理; 文本图像; 二值化; 图分割; 谱聚类
Abstract: The located text regions need to be binarized for accurate recognition in automatic textural extraction. Due to the complex backgrounds, traditional thresholding methods can not segment the character image effectively from natural scenes. A novel approach of binarization is proposed for gray images. The proposed algorithm uses the Normalized graph cut(Ncut) as the measure for spectral clustering, and the weighted matrices used in evaluating the graph cuts are based on the gray levels of an image, rather than the commonly used image pixels. Thus, the proposed algorithm requires much smaller spatial costs and much lower computation complexity. Experiments on text images in natural scene show the superior performance of the proposed method compared to the typical thresholding algorithms.

HTML全文

参考文献(1)

Lienhart R and Wernicke A. Localizing and segmenting textin images and videos[J].IEEE Transactions on Circuits andSystems for Video Technology.2002, 12(4):256-268[2]Mariano V Y and Kasturi R. Locating uniform-colored textin video frames[J].Proc. of Intl Conference on PatternRecognition, Barcelona, Spain.2000, 4:539-542[3]Chen D, Odobez J M, and Bourlard H. Text detection andrecognition in images and video frames[J].Pattern Recognition.2004, 37(3):595-608[4]Zhong Yu, Zhang Hong-jiang, and Jain A K. Automaticcaption localization in compressed video[J].IEEE Transactionson Pattern Analysis and Machine Intelligence.2000, 22(4):385-392[5]Chen Xi-lin, Yang Jie, Zhang Jing, and Waibel A. Automaticdetection and recognition of signs from natural scenes[J].IEEETransactions on Image Processing.2004, 13(1):87-99[6]Chen Xiang-rong and Yuille A L. Detecting and reading textin natural scenes. Proceedings of the IEEE Computer SocietyConference on Computer Vision and Pattern Recognition,Washington, DC, USA, 2004: 366-373.[7]Tsai T H and Chen Y C. A comprehensive motion videotextdetection localization and extraction method. Proc. of IEEEIntl Conference on Data Engineering Workshop, Istanbul,Turkey, 2007: 113-116.Pan W M.[J].Bui T D, and Suen CY. Text segmentation fromcomplex background using sparse representations. Proc.ofIntl Conference on Document Analysis Recognition, Curitiba,Brazil.2007,:-[8]Wu V, Manmatha R, and Riseman E M. Text finder: Anautomatic system to detect and recognize text in images[J].IEEE Transactions on Pattern Analysis and MachineIntelligence.1999, 21(11):1224-1229[9]Otsu N. A threshold selection method from grey levelhistograms[J].IEEE Transactions on Systems, Man andCybernetics.1979, 9(1):62-66[10]Wu Z Y and Leahy R. An optimal graph theoretic approachto data clustering: Theory and its application to imagesegmentation[J].IEEE Transactions on Pattern Analysis andMachine Intelligence.1993, 15(11):1101-1113[11]Shi Jian-bo and Malik J. Normalized cuts and imagesegmentation[J].IEEE Transactions on Pattern Analysis andMachine Intelligence.2000, 22(8):888-905[12]He X, Cai D, and Wen J R, et al.. Clustering and searchingWWW images using link and page layout analysis. ACMTransactions on Multimedia Computing, Communicationsand Applications, 2007, 3(2): Article No. 10.[13]Higham D J, Kalna G, and Kibble M. Spectral clustering andits use in bioinformatics[J].Journal of Computational andApplied Mathematics.2007, 204(1):25-37[14]陶文兵, 金海. 一种新的基于图谱理论的图像阈值分割方法.计算机学报, 2007, 30(1): 110-118.Tao Wen-bing and Jin Hai. A new image thresholding methodbased on graph spectral theory. Chinese Journal ofComputers, 2007, 30(1): 110-118.[15]Lucas S M.[J].Panaretos A, and Sosa L, et al.. ICDAR 2003robust reading competition. Proc. of 7th Intl Conference onDocument Analysis and Recognition, Scotland.2003,:-

施引文献

资源附件(0)

访问统计