Methods for objective evaluation and improvement of text document images

Valery S. Kot; Alexander V. Bondarenko

doi:10.1117/12.323158

1 October 1998 Methods for objective evaluation and improvement of text document images

Valery S. Kot, Alexander V. Bondarenko

Proceedings Volume 3460, Applications of Digital Image Processing XXI; (1998) https://doi.org/10.1117/12.323158
Event: SPIE's International Symposium on Optical Science, Engineering, and Instrumentation, 1998, San Diego, CA, United States

Abstract

Any optical character recognition (OCR) system contains preprocessing unit responsible for image binarization, and the entire recognition rate depends dramatically from the accuracy of this unit. In case of poor image quality user must spend much time to find out the best parameters of this unit while recognition rate may still remain unsatisfactory. Thus methods intended for objective evaluation and context- sensitive improvement of text document images are required. In this parameters set is proposed as a tool for integral image description. This compact set allows to select automatically or semiautomatically the optimal image processing sequence from the basic IP functions. For all tested commercial OCR systems, the proposed methods result in recognition errors decreasing about 50 - 60% for text document images of average and poor quality while requiring less than 1 minute per page of additional processing time.

Citation Download Citation

Valery S. Kot and Alexander V. Bondarenko "Methods for objective evaluation and improvement of text document images", Proc. SPIE 3460, Applications of Digital Image Processing XXI, (1 October 1998); https://doi.org/10.1117/12.323158

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