PSO-based stochastic resonance for automatic contrast enhancement of images

Images suffer from poor contrast due to insufficient illumination. The dynamic range of pixel intensities in low-contrast images is very limited, thus not revealing the complete details in the image. Contrast enhancement techniques are used to improve the visual quality of such low-contrast images....

Full description

Saved in:
Bibliographic Details
Published in:Signal, image and video processing image and video processing, 2016-01, Vol.10 (1), p.207-214
Main Authors: Maragatham, G., Roomi, S. Md. Mansoor
Format: Article
Language:English
Subjects:
Computer Imaging
Computer Science
Image Processing and Computer Vision
Multimedia Information Systems
Original Paper
Pattern Recognition and Graphics
Signal,Image and Speech Processing
Vision
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Images suffer from poor contrast due to insufficient illumination. The dynamic range of pixel intensities in low-contrast images is very limited, thus not revealing the complete details in the image. Contrast enhancement techniques are used to improve the visual quality of such low-contrast images. Many blind algorithms have been developed to enhance the image without understanding whether it requires such an enhancement procedure. In this paper, we present an automatic contrast enhancement method where the need of the contrast enhancement is ascertained by modelling image contrast factor and using log-likelihood function. Once the need is identified, particle swarm optimization-based stochastic resonance is applied to improve the contrast of the image. The proposed method significantly improves the contrast and simultaneously preserves the brightness of the low-contrast images without introducing any artefacts. Quantitative evaluation of a comparative experimentation with a few conventional enhancement methods demonstrates that the proposed method achieves better-quality results.
ISSN:1863-1703
1863-1711
DOI:10.1007/s11760-014-0728-2