Image Quality Assessment Based on Sparse Neighbor Importance

Abstract

In this paper, the image quality assessment problem is discussed from the perspective of sparse coding and a new automatic image quality assessment algorithm is presented. Specifically, the input image is first divided into non-overlapping blocks and sparse coding is used to reconstruct a central sub-block using neighboring sub-blocks as dictionaries. Two-dimensional sparse vectors from each neighboring sub-block are devised as importance maps, which are then used in similarity measurements between the reference and distorted images. The proposed method was compared with several recently introduced shallow and deep methods across four datasets and multiple distortion types. The experimental results obtained show that it has a strong correlation with the Human Visual System and outperforms its counterparts. © 2022 Elsevier B.V., All rights reserved.