Premium
Contrast Improvement Method for Monochrome Images Based on the Ratio of Local Variances of Gray levels
Author(s) -
Saitoh Fumihiko
Publication year - 2008
Publication title -
ieej transactions on electrical and electronic engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.254
H-Index - 30
eISSN - 1931-4981
pISSN - 1931-4973
DOI - 10.1002/tee.20273
Subject(s) - monochrome , contrast (vision) , histogram equalization , artificial intelligence , pixel , histogram , brightness , gray level , computer vision , contrast ratio , gray (unit) , computer science , mathematics , image contrast , pattern recognition (psychology) , image (mathematics) , optics , physics , medicine , radiology , operating system
When the contrast of a captured image is not good, a method for contrast improvement may be useful. Especially, liner transform (LT) of gray levels and the histogram equalization (HE) method are well known as general methods. However, an effective contrast improvement does not always work, or an unnatural output is often generated by these methods. This paper proposes a method for contrast improvement to generate an image with medium contrast between those obtained by LT and HE. An input image is separated into an object area and a background area by every pixel. The variance of gray levels in the local area is used for measuring the local contrast. The ratio of the variances of gray levels in these areas is evaluated to determine the relative weights of LT and HE. The experimental results show that output images with higher contrast than that by LT and with less saturation of brightness and less noise than those by HE were generated by the proposed method. Copyright © 2008 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.