Share Email Print

Proceedings Paper

Perceptual color noise formulation
Author(s): Jiangtao Kuang; Xiaoyun Jiang; Shuxue Quan; Andrew Chiu
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Image noise is one of the important image quality metrics for evaluation and optimization in color reproduction system. While the noise evaluation in psychometric lightness L* works pretty well in black-and-white images, it is insufficient for color images. A perceptual color noise evaluation equation was derived to extend noise evaluation in CIELAB color space, incorporating chromatic noise components. Psychophysical experiments were designed to evaluate color noise subjectively. In the first step, Weber's Law and Fechner's Law were used to generate a standard ruler with equal perceptual interval steps, which served to anchor a numerical image quality rating scale for the subjective rating experiment in the second step. An objective noise evaluation equation that incorporated the noise sensitivity functions modeled from the experimental data was found to better correlate with subjective evaluation. A more robust noise evaluation equation will be derived based on the psychophysical experiment techniques and experimental data in the future.

Paper Details

Date Published: 17 January 2005
PDF: 8 pages
Proc. SPIE 5668, Image Quality and System Performance II, (17 January 2005); doi: 10.1117/12.587807
Show Author Affiliations
Jiangtao Kuang, Rochester Institute of Technology (United States)
Xiaoyun Jiang, Qualcomm Inc. (United States)
Shuxue Quan, Qualcomm Inc. (United States)
Andrew Chiu, Qualcomm Inc. (United States)

Published in SPIE Proceedings Vol. 5668:
Image Quality and System Performance II
Rene Rasmussen; Yoichi Miyake, Editor(s)

© SPIE. Terms of Use
Back to Top