Name
Title | ||
|---|---|---|
Quality assessment of images with multiple distortions based on phase congruency and gradient magnitude. |
Abstract | ||
|---|---|---|
In image communication systems, images are often contaminated by multiple types of distortions. However, most existing image quality assessment (IQA) methods mainly focused on a single type of distortions. In this paper, we proposed a no-reference (NR) IQA method for images with multiple distortions. Image distortions not only destroy the intensity of low-level image features, but also alter their distribution, to both of which the human vision system (HVS) is sensitive. Based on these observations, low-level features are represented by phase congruency (PC) which is consistent with human perception. The distribution of low-level features is extracted using local binary pattern (LBP) in PC domain at multiple scales, which can effectively characterize the impact of multiple distortions on images. Given that PC is contrast invariant while the contrast does affect perceptual image quality of the HVS, image gradient magnitude (GM) is employed as a weighting factor for LBP histogram creation. Finally a support vector regression model is trained to map the gradient-weighted LBP histograms in PC domain at multi-scale to quality scores. Experimental results on two benchmark databases demonstrate that the proposed method achieves high consistency with subjective perception and performs better than other state-of-the-art full-reference (FF) and NR IQA methods. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1016/j.image.2019.08.013 | Signal Processing: Image Communication |
Keywords | Field | DocType |
No-reference,Image quality assessment,Phase congruency,Local binary pattern,Image gradient | Computer vision,Histogram,Image gradient,Machine vision,Feature (computer vision),Computer science,Support vector machine,Local binary patterns,Image quality,Artificial intelligence,Phase congruency | Journal |
Volume | ISSN | Citations |
79 | 0923-5965 | 0 |
PageRank | References | Authors |
0.34 | 0 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Xi-kui Miao | 1 | 0 | 0.34 |
| Hairong Chu | 2 | 5 | 1.93 |
| Hui Liu | 3 | 0 | 0.68 |
| Yao Yang | 4 | 0 | 0.34 |
| Xiaolong Li | 5 | 0 | 0.34 |