A Contrast Enhancement Method for HDR Image Using a Modified Image Formation Model

Byoung-Ju YUN  Hee-Dong HONG  Ho-Hyoung CHOI  

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E95-D   No.4   pp.1112-1119
Publication Date: 2012/04/01
Online ISSN: 1745-1361
DOI: 10.1587/transinf.E95.D.1112
Print ISSN: 0916-8532
Type of Manuscript: PAPER
Category: Image Processing and Video Processing
Keyword: 
modified image formation model,  global illumination,  local illumination,  reflectance,  JND-based adaptive smoothing method,  

Full Text: PDF(1.2MB)>>
Buy this Article




Summary: 
Poor illumination and viewing conditions have negativeinfluences on the quality of an image, especially the contrast of the dark and bright region. Thus, captured and displayed images usually need contrast enhancement. Histogram-based or gamma correction-based methods are generally utilized for this. However, these methods are global contrast enhancement method, and since the sensitivity of the human eye changes locally according to the position of the object and the illumination in the scene, the global contrast enhancement methods have a limit. The spatial adaptive method is needed to overcome these limitations and it has led to the development of an integrated surround retinex (ISR), and estimation of dominant chromaticity (EDC) methods. However, these methods are based on Gray-World Assumption, and they use a general image formation model, so the color constancy is known to get poor results, shown through graying-out, halo-artifacts (ringing effects), and the dominated color. This paper presents a contrast enhancement method using a modified image formation model in which the image is divided into three components: global illumination, local illumination and reflectance. After applying the power constant value to control the contrast in the resulting image, the output image is obtained from their product to avoid or minimize a color distortion, based on the sRGB color representation. The experimental results show that the proposed method yields better performances than conventional methods.