Evaluating image quality using consistent grey relational grade

Abstract

Purpose

Assessing image quality is a difficult task. Different demands need distinct criteria, so it is not realistic to decide which contrast enhancement method is better only through one criterion. The main purpose is to propose an efficient scheme to effectively evaluate image quality. Furthermore, the idea can be applied in other fields.

Design/methodology/approach

To objectively and quantitatively assess image quality, the authors integrate four criteria into one composite criterion and use it to evaluate seven existing contrast enhancement methods. The mechanism of integration is through a newly proposed way of computing a grey relational grade (GRGd), called the consistent grey relational grade (CGRGd).

Findings

In this paper, the authors propose the CGRGd, which is more efficient and consistent than other existing GRGds. When applied to image quality evaluation, the proposed CGRGd can effectively choose the best method than others. The results also indicate that the proposed CGRGd combined with appropriate criteria can be widely used in the field of multiple criteria.

Originality/value

The proposed CGRGd is a new approach to the problem of multi-criteria evaluation, and its application to the evaluation of image quality is a novel idea. For readers interested in the field of multi-criteria decision-making, the CGRGd provides an efficient and effective alternative.

Citation

Chang, Y.-C., Chang, C.-M., Chen, L.-H. and Chan, T.-J. (2014), "Evaluating image quality using consistent grey relational grade", Engineering Computations, Vol. 31 No. 2, pp. 231-249. https://doi.org/10.1108/EC-01-2013-0016