《计算机应用研究》|Application Research of Computers

一种自适应烟标字符提取方法

Method of adaptive cigarette mark character extraction

免费全文下载 (已被下载 次)  
获取PDF全文
作者 殷羽,郑宏,王静,李圣,王震
机构 武汉大学 电子信息学院,武汉 430072
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2015)04-1248-05
DOI 10.3969/j.issn.1001-3695.2015.04.069
摘要 针对印刷质量不高、背景复杂的烟标字符难以提取问题,提出了一种自适应的烟标字符提取算法。通过对单通道烟标图像进行背景重构后,根据支持向量机中得到的训练模型自适应选取最优二值化方法;提取标准轮廓定位字符区域;结合标准轮廓,同时根据最优二值化图像投影直方图进行烟标字符提取。经过实验验证,该方法能够自适应地选择最优二值化方法,字符提取精度可达到90%以上,同时原理简单、鲁棒性强并便于硬件实现。
关键词 支持向量机;背景重构;N-Otsu法;最优二值化;字符提取
基金项目 高等学校博士学科点专项科研基金资助项目(20110141110044)
本文URL http://www.arocmag.com/article/01-2015-04-069.html
英文标题 Method of adaptive cigarette mark character extraction
作者英文名 YIN Yu, ZHENG Hong, WANG Jing, LI Sheng, WANG Zhen
机构英文名 School of Electronic Information, Wuhan University, Wuhan 430072, China
英文摘要 This paper proposed a cigarette mark character extraction method which could adaptively solve the problem that cigarette image difficult to extract.By background reconstruction to the single channel images, the method adaptively selected the optimal binarization based on SVM, and extracted the standard contours for the character areas.According to the standard contours and based on the best binary image histogram projection, the method extacted the cigarette characters.After experimental verification, it adaptively selected the best binarization and character extraction accuracy could reach 90%.Simultaneously the method has simple principles and strong robustness and convenient for hardware implementation.
英文关键词 SVM; background reconstruction; N-Otsu method; optimal binarization; character extraction
参考文献 查看稿件参考文献
  [1] EPSHTEIN B, OFEK E, WEXLER Y. Detecting text in natural scenes with stroke width transform[C] //Proc of IEEE Conference on Computer Vision and Pattern Recognition. [S. l. ] :IEEE Press, 2010:2963-2970.
[2] LIU Xiao-qing, SAMARABANDU J. Multiscale edge-based text extraction from complex images[C] //Proc of IEEE International Conference on Multimedia and Expo. [S. l. ] :IEEE Press, 2006:1721-1724.
[3] DU Yu-ning, DUAN Gen-quan, AI Hai-zhou. Context-based text detection in natural scenes[C] //Proc of the 19th IEEE International Conference on Image Processing. [S. l. ] :IEEE Press, 2012:1857-1860.
[4] YI Chu-cai, TIAN Ying-li. Text detection in natural scene images by stroke Gabor words[C] //Proc of the 12th International Conference on Document Analysis and Recognition. [S. l. ] :IEEE Press, 2011:177-181.
[5] PHAN T Q, SHIVAKUMARA P, TAN C L. A Laplacian method for video text detection[C] //Proc of the 10th International Conference on Document Analysis and Recognition. [S. l. ] :IEEE Press, 2009:66-70.
[6] ZHOU Zhi-wei, LI Lin-lin, TAN C L. Edge based binarization for video text images[C] //Proc of the 20th International Conference on Pattern Recognition. [S. l. ] :IEEE Press, 2010:133-136.
[7] ANTHIMOPOULOS M, GATOS B, PRATIKAKIS I, et al. Detecting text in video frames[C] //Proc of the 4th International Conference on Signal Processing, Pattern Recognition and Applications. 2007:40-44.
[8] AGHAJARI G, SHANBEHZADEH J, SARRAFZADEH A. A text localization algorithm in color image via new projection profile[C] //Proc of International Multiconference of Engineers and Computer Scientists. 2010.
[9] 樊汝策, 王庆, 翟正军, 等. 一种改进的针对退化文本图像的二值化方法[J] . 测控技术, 2013, 32(5):29-31.
[10] 王霞玲, 吕岳, 文颖. 复杂背景和非均匀光照环境下的条码自动定位和识别[J] . 智能系统学报, 2010, 5(1):35-40.
[11] 吴祖堂, 朱玉荣, 黄晓飞. Otsu阈值改进及其在俯仰角测试中的应用[J] . 弹道学报, 2012, 24(1):107-110.
[12] CHOI J H, LEE H Y, LEE H K. Color laser printer forensic based on noisy feature and support vector machine classifier[J] . Multimedia Tools and Applications, 2013, 67(2):363-382.
[13] 代小红. 模糊模式的手写数字识别技术研究与实现[J] . 重庆大学学报:自然科学版, 2011, 34(6):117-122.
[14] SAUVOLA J, PIETIKINEN M. Adaptive document image binarization[J] . Pattern Recognition, 2000, 33(2):225-236.
[15] 申森, 李艾华, 姚良, 等. 基于小波包和 Niblack 法的枪号图像二值化算法[J] . 光子学报, 2013, 42(3):354-358.
[16] 陈琪, 熊博莅, 陆军, 等. 改进的二维Otsu图像分割方法及其快速实现[J] . 电子与信息学报, 2010, 32(5):1100-1104.
[17] 吴锐, 黄剑华, 唐降龙, 等. 基于灰度直方图和谱聚类的文本图像二值化方法[J] . 电子与信息学报, 2009, 31(10):2460-2464.
收稿日期 2014/3/6
修回日期 2014/4/24
页码 1248-1252
中图分类号 TP391.43
文献标志码 A