AI helps you reading Science

AI generates interpretation videos

AI extracts and analyses the key points of the paper to generate videos automatically


pub
Go Generating

AI Traceability

AI parses the academic lineage of this thesis


Master Reading Tree
Generate MRT

AI Insight

AI extracts a summary of this paper


Weibo:
We have proposed a general-purpose and accurate method based upon bagging for skew estimation of document images

Skew estimation of document images using bagging.

IEEE Transactions on Image Processing, no. 7 (2010): 1837-1846

Cited: 30|Views81
WOS EI SCOPUS

Abstract

This paper proposes a general-purpose method for estimating the skew angles of document images. Rather than to derive a skew angle merely from text lines, the proposed method exploits various types of visual cues of image skew available in local image regions. The visual cues are extracted by Radon transform and then outliers of them are ...More

Code:

Data:

0
Introduction
  • N OWADAYS, the scanner is becoming a common device in modern office for converting paper document into its digital format.
  • Document skew commonly occurs during the scanning process when paper documents are not fed properly into the scanner.
  • The existence of such artifacts will cause some significant problems in the subsequent steps, such as page layout analysis and character recognition.
  • The basic principle is to derive a skew angle by exploring certain visual cues.
  • Proposed — including the projection profiles based techniques [5]–[9], the transition counts based techniques [10], [11], the cross-correlation based techniques [12], [13], the Hough
Highlights
  • N OWADAYS, the scanner is becoming a common device in modern office for converting paper document into its digital format
  • We propose a general-purpose method in this paper for skew estimation by integrating multiple local visual cues of image skew using a bagging estimator
  • The second one is performed on the 1600 images scanned from real printed English journals. The goal of this experiment is to test the robustness of the proposed method to noises, spare and short text lines, and the presence of large areas of nontextual objects etc
  • We have proposed a general-purpose and accurate method based upon bagging for skew estimation of document images
  • The visual cues are first extracted by Radon transform and the outliers of them are iteratively rejected through a floating cascade
  • To evaluate the performance of the proposed method, a series of large-scale experiments based upon UWDB-III are conducted
Results
  • To evaluate the performance of the proposed method, three experiments are implemented on the University of Washington

    English Document Image Database III (UWDB-III).
  • The skew angles of these images are exactly 0 , so they are quite suitable for testing the correctness and accuracy of the proposed method.
  • The second one is performed on the 1600 images scanned from real printed English journals.
  • The goal of this experiment is to test the robustness of the proposed method to noises, spare and short text lines, and the presence of large areas of nontextual objects etc.
  • The third one is a comparative experiment on UWDB-I, which aims to make a comparison of the method with the state-of-the-art methods that were evaluated on the same dataset
Conclusion
  • As the noise density increases, visual cues of image skew are ruined and difficult to be robustly extracted.In this paper, the authors have proposed a general-purpose and accurate method based upon bagging for skew estimation of document images.
  • The experimental results show that the proposed method is an accurate and general-purpose method
  • It is highly competitive in execution speed and estimation accuracy, and extremely robust to document noises, multiple different skews, short and sparse text lines, and the presence of large areas of nontextual objects of various types and quantities
Tables
  • Table1: PERFORMANCE COMPARISON OF FIVE METHODS USING
  • Table2: AVERAGE ABSOLUTE ERRORS (AAE) AND STANDARD DEVIATIONS (S. D.) OF SYNTHETIC IMAGES WITH DIFFERENT FONT SIZES
  • Table3: SKEW ESTIMATION ERRORS OF IMAGE L01DSYN
  • Table4: AVERAGE ABSOLUTE ERRORS (AAE) AND STANDARD DEVIATIONS (S. D.) OF
  • Table5: SKEW ESTIMATION ERRORS OF IMAGE L01DSYN (SKEW ANGLE: 10.420 )
Download tables as Excel
Funding
  • This work was supported in part by the National Science Foundation of China under grant No 60675012, 60635050, and in part by the National “863” HighTech Program of China under grant No 2009AA012104
Reference
  • Y. Y. Tang, S. W. Lee, and C. Y. Suen, “Automatic document processing: A survey,” Pattern Recognit., vol. 29, no. 12, pp. 1931–1952, 1996.
    Google ScholarLocate open access versionFindings
  • G. Nagy, “Twenty years of document image analysis in PAMI,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 22, no. 1, pp. 38–62, Jan. 2000.
    Google ScholarLocate open access versionFindings
  • A. D. Bagdanov and J. Kanai, “Evaluation of document image skew estimation techniques,” in Proc. SPIE, San Jose, CA, 1996, vol. 2660, pp. 343–353.
    Google ScholarLocate open access versionFindings
  • J. J. Hull, “Document image skew detection: Survey and annotated bibliography,” in Document Analysis Systems II, J. J. Hull and S. L. Taylor, Eds. Singapore: World Scientific, 1998, pp. 40–64.
    Google ScholarFindings
  • H. S. Baird, “The skew angle of printed documents,” in Proc. SPSE 40th Ann. Conf. and Symp. on Hybrid Imaging Systems, Rochester, NY, 1987, pp. 21–24.
    Google ScholarFindings
  • E. Kavallieratou, N. Fakotakis, and G. Kokkinakis, “Skew angle estimation for printed and handwritten documents using the wigner-ville distribution,” Image Vis. Comput., vol. 20, no. 11, pp. 813–824, 2002.
    Google ScholarLocate open access versionFindings
  • N. Liolios, N. Fakotakis, and G. Kokkinakis, “On the generalization of the form identification and skew detection problem,” Pattern Recognit., vol. 35, no. 1, pp. 253–264, 2002.
    Google ScholarLocate open access versionFindings
  • R. Kapoor, D. Bagai, and T. S. Kamal, “A new algorithm for skew detection and correction,” Pattern Recognit. Lett., vol. 25, no. 11, pp. 1215–1229, 2004.
    Google ScholarLocate open access versionFindings
  • S. T. Li, Q. H. Shen, and J. Sun, “Skew detection using wavelet decomposition and projection profile analysis,” Pattern Recognit. Lett., vol. 28, no. 5, pp. 555–562, 2007.
    Google ScholarLocate open access versionFindings
  • Y. Ishitani, “Document skew detection based on local region complexity,” in Proc. 2nd Int. Conf. Document Analysis and Recognition, Tsukuba, Japan, 1993, pp. 49–52.
    Google ScholarLocate open access versionFindings
  • Y. K. Chen and J. F. Wang, “Skew detection and reconstruction based on maximization of variance of transition-counts,” Pattern Recognit., vol. 33, no. 2, pp. 195–208, 2000.
    Google ScholarLocate open access versionFindings
  • H. Yan, “Skew correction of document images using interline crosscorrelation,” CVGIP, Graph. Models Image Process., vol. 55, no. 6, pp. 538–543, 1993.
    Google ScholarLocate open access versionFindings
  • B. Gatos, N. Papamarkos, and C. Chamzas, “Skew detection and text line position determination in digitized documents,” Pattern Recognit., vol. 30, no. 9, pp. 1505–1519, 1997.
    Google ScholarLocate open access versionFindings
  • D. S. Le, G. R. Thoma, and H. Wechsler, “Automated page orientation and skew angle detection for binary document images,” Pattern Recognit., vol. 27, no. 10, pp. 1325–1344, 1994.
    Google ScholarLocate open access versionFindings
  • B. Yu and A. K. Jain, “A robust and fast skew detection algorithm for generic documents,” Pattern Recognit., vol. 29, no. 10, pp. 1599–1629, 1996.
    Google ScholarLocate open access versionFindings
  • A. Amin and S. Fischer, “A document skew detection method using the hough transform,” Pattern Anal. Appl., vol. 3, no. 3, pp. 243–253, 2000.
    Google ScholarLocate open access versionFindings
  • H. K. Kwag, S. H. Kim, S. H. Jeong, and G. S. Lee, “Efficient skew estimation and correction algorithm for document images,” Image Vis. Comput., vol. 20, no. 1, pp. 25–35, 2002.
    Google ScholarLocate open access versionFindings
  • C. Singh, N. Bhatia, and A. Kaur, “Hough transform based fast skew detection and accurate skew correction methods,” Pattern Recognit., vol. 41, no. 12, pp. 3528–3546, 2008.
    Google ScholarLocate open access versionFindings
  • L. O’Gorman, “The document spectrum for page layout analysis,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 15, no. 11, pp. 1162–1173, Nov. 1993.
    Google ScholarLocate open access versionFindings
  • Y. Lu and C. L. Tan, “A nearest-neighbor chain based approach to skew estimation in document images,” Pattern Recognit. Lett., vol. 24, no. 14, pp. 2315–2323, 2003.
    Google ScholarLocate open access versionFindings
  • B. Yuan and C. L. Tan, “Fiducial line based skew estimation,” Pattern Recognit., vol. 38, no. 12, pp. 2333–2350, 2005.
    Google ScholarLocate open access versionFindings
  • V. Aradhya, G. H. Kumar, and P. Shivakumara, “An accurate and efficient skew estimation technique for south indian documents: A new boundary growing and nearest neighbor clustering based approach,” Int. J. Robot. Autom., vol. 22, no. 4, pp. 272–280, 2007.
    Google ScholarLocate open access versionFindings
  • H. Liu, Q. Wu, H. B. Zha, and X. P. Liu, “Skew detection for complex document images using robust borderlines in both text and non-text regions,” Pattern Recognit. Lett., vol. 29, no. 13, pp. 1893–1900, 2008.
    Google ScholarLocate open access versionFindings
  • C. H. Chou, S. Y. Chu, and F. Chang, “Estimation of skew angles for scanned documents based on piecewise covering by parallelograms,” Pattern Recognit., vol. 40, no. 2, pp. 443–455, 2007.
    Google ScholarLocate open access versionFindings
  • B. Yuan and C. L. Tan, “Convex hull based skew estimation,” Pattern Recognit., vol. 40, no. 2, pp. 456–475, 2007.
    Google ScholarLocate open access versionFindings
  • L. Breiman, “Bagging predictors,” Mach. Learn., vol. 24, no. 2, pp. 123–140, 1996.
    Google ScholarLocate open access versionFindings
  • J. V. Hansen, “Combining predictors: Comparison of five meta machine learning methods,” Inf. Sci., vol. 119, no. 1–2, pp. 91–105, 1999.
    Google ScholarLocate open access versionFindings
  • Avanindra and S. Chaudhuri, “Robust detection of skew in document images,” IEEE Trans. Image Process., vol. 6, no. 2, pp. 344–349, Feb. 1997.
    Google ScholarLocate open access versionFindings
  • S. R. Deans, The Radon Transform and Some of Its Applications. New York: Wiley, 1983.
    Google ScholarFindings
  • S. Chen, “Document Layout Analysis Using Recursive Morphological Transforms,” Ph.D. dissertation, Univ. Washington, Seattle, WA, 1995.
    Google ScholarFindings
  • D. S. Bloomberg, G. E. Kopec, and L. Dasari, “Measuring document image skew and orientation,” in Proc. SPIE, Document Recognition II, San Jose, CA, 1995, vol. 2422, pp. 302–316.
    Google ScholarLocate open access versionFindings
  • P. Shivakumara and G. H. Kumar, “A novel boundary growing approach for accurate skew estimation of binary document images,” Pattern Recognit. Lett., vol. 27, no. 7, pp. 791–801, 2006. Gaofeng Meng received the B.S. degree in applied mathematics from Northwestern Polytechnical University, Xi’an, China, in 2002, and the M.S. degree in applied mathematics from Tianjin University, Tianjin, China, in 2005, and the Ph.D. degree in control science and engineering from Xi’an Jiaotong University, Xi’an, Shaanxi, China, in 2009.
    Google ScholarLocate open access versionFindings
  • In 2009, he jointed the National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing, China as an Assistant Professor. His current research interests include image processing, computer vision, and pattern recognition.
    Google ScholarFindings
  • Chunhong Pan received the B.S. degree in automatic control from Tsinghua University, Beijing, China, in 1987, the M.S. degree from Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, China, in 1990, and the Ph.D. degree in pattern recognition and intelligent system from Institute of Automation, Chinese Academy of Sciences, Beijing, in 2000.
    Google ScholarLocate open access versionFindings
  • Nanning Zheng (SM’94–F’06) received the B.S. degree from the Department of Electrical Engineering, Xi’an Jiaotong University, Xi’an, China, in 1975, the M.S. degree in information and control engineering from Xi’an Jiaotong University in 1981, and the Ph.D. degree in electrical engineering from Keio University, Yokohama, Japan, in 1985.
    Google ScholarLocate open access versionFindings
  • Chen Sun received the B.S. degree in information engineering and the M.S. degree in control science and engineering from Xi’an Jiaotong University in 2005 and 2008, respectively, and the general engineering degree from Ecole Centrale Paris in 2006. In July 2008, she joined Shanghai Baosight Software Co., Ltd. as a software engineer. Her current research interests include image processing and pattern recognition.
    Google ScholarFindings
0
Your rating :

No Ratings

Tags
Comments
数据免责声明
页面数据均来自互联网公开来源、合作出版商和通过AI技术自动分析结果,我们不对页面数据的有效性、准确性、正确性、可靠性、完整性和及时性做出任何承诺和保证。若有疑问,可以通过电子邮件方式联系我们:report@aminer.cn