近来,文档图像的计算机自动理解已取得很多进展。但是,对于具有倾斜的图像的理解仍然存在许多困难。这种困难在中文名片图像自动识别与理解系统中尤为突出。必须在系统的输入端对图像作有效的倾斜校正以保证系统的性能。由于中文名片版面复杂,名片中文字行以及每行字符较少,使得现有的倾斜校正算法在处理名片图像时效果很不理想。Hough变换可用于一般文档图像的倾斜校正。但是,Hough变换在名片图像中的应用还有待研究。本文提出一种二级Hough变换算法,并应用于名片图像理解系统,利用名片图像自身的特点提高Hough变换的精确度和速度。这一方法的效果已被实验结果所证实。
Abstract
Automatic document understanding has undergone great progresses in the past decade. Yet the difficulties due to image skewness have not been overcome. Such difficulties are especially vital in Chinese Business Card understanding systems.Because of the complex card layout ,none of current de-skew algorithms shows satisfactory performance on Chinese Business Card image. Although Hough transform has been widely used in the de-skew of general document image ,it’s application to Chinese Business Card image needs more research effort . In this paper ,we proposed a two-stage Hough transform algorithm and applied it to our card understanding system. This algorithm takes advantage of the characteristics of card images and improves both accuracy and time complexity. Such improvement has been proven by the experimental results.
关键词
文档分析 /
版面理解 /
倾斜校正 /
Hough变换 /
中文名片
{{custom_keyword}} /
Key words
document analysis /
layout understanding /
deskew detection /
hough transform
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Chiu L Yu , Yuan Y Tang ,Ching Y Suen. Document Skew Detection Based on the Fractal and Least Squares Method. Proc. 3rd ICDAR ,1995 ,1149 - 1152
[2] Ray Smith. A Simple and Efficient Skew Detection Algorithm via Text Row Accumulation. Proc. 3rd ICDAR ,1995 ,1145 - 1148
[3] Bagdanov A ,Kanai J . Projection Profile Based Skew Estimation Algorithm for JBIG Compressed Images. Proc. 4th ICDAR ,1997 ,401 - 405
[4] Akiyama T ,Hagita N. Automated Entry System for Printed Documents. Pattern Recognition ,1990 ,23 (11) :1141 - 1154
[5] Postl W. Detection of Linear Oblique Structures and Skew Scan in Digitized Documents. Proc. 8th ICPR , Paris ,France ,1986 ,687 - 689
[6] Hashizume A ,Yeh P-S ,Rosenfeld A. A method of detecting the orientation of aligned components. Pattern Recognition Letters ,1986 ,4 :125 - 132
[7] O’Gorman L. The Document Spectrum for Page Layout Analysis. IEEE Trans. on PAMI ,1993 ,15 (11) : 1162 - 1173
[8] Hinds S C ,Fisher J L ,D’Amato D P. A document skew detection method using run-length encoding and the Hough transform In :Proceedings of International Conference on Pattern Recognition ,1990 , I :464 - 468
[9] Duda R O ,Hart P E. Use of Hough Transform to Detect Lines and Curves in Pictures : Graphics and Image Processing. Comm. ACM ,1972 ,15 :11 - 15
[10] Illingworth J ,Kittler J . A Survey of the Hough Transform. Computer Vision ,Graphics ,and Image Processing ,1998 ,44 ,87 - 116
[11] Illingworth J , Kittler J . The adaptive Hough Transform. IEEE Trans. Pattern Anal ,Mach. Intell. PAMI ,1987 ,9 (5) :691 - 698
[12] Li H ,Lavin M A ,LeMaster R J . Fast Hough Transform. Research Report RC11080 (# 49754) , IBM T. J . Watson Research Center ,P. O.Box 218 ,Yorktown Heights ,NY,March 1985
[13] Thomas Risse. Hough Transform for Line Recognition :Complexity of Evidence Accumulation and Cluster Detection. Computer Vision ,Graphics and Image Processing ,1989 ,46 :327 - 345
[14] Wilson C Y Lam ,Lam T S Lam ,Kelvin S Y Yuen et al . An Analysis on Quantizing the Hough Space. Pattern Recognition Letters ,1994 ,15 ,1127 - 1135
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(269 KB)
