Geometrical approach to skew detection for documents containing the Latin/Cyrillic characters

Oleg G. Okun

doi:10.1117/12.364111

23 September 1999 Geometrical approach to skew detection for documents containing the Latin/Cyrillic characters

Oleg G. Okun

Proceedings Volume 3811, Vision Geometry VIII; (1999) https://doi.org/10.1117/12.364111
Event: SPIE's International Symposium on Optical Science, Engineering, and Instrumentation, 1999, Denver, CO, United States

Abstract

Document skew is a distortion mainly concerning the orientation of text lines and occurring when digitizing the paper documents. Its visual effect is a slope of text lines, which are normally horizontal for such scripts as Latin or Cyrillic, with respect to the X-axis. Many available document recognition systems, however, require properly aligned text liens for accurate text segmentation and recognition. It means that the skew, if present, should be estimated and compensated before further processing. The Hough transform is one of the popular techniques for skew detection. To lower its computational cost, it is usually applied to a small number of representative points of each character or its bounding box. However, a problem with this method is that different characters have different heights. As a result, the representative points of characters belonging to the same line often do not fit well to a straight line and this often leads to errors in skew detection by using the Hough transform. In this paper, we propose a new algorithm to overcome this problem. It only uses the bounding boxes of the connected components of characters and a number of simple tests in order to obtain the skew angle estimation.

Citation Download Citation

Oleg G. Okun "Geometrical approach to skew detection for documents containing the Latin/Cyrillic characters", Proc. SPIE 3811, Vision Geometry VIII, (23 September 1999); https://doi.org/10.1117/12.364111

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available