Full-reference image quality assessment by combining features in spatial and frequency domains
Objective image quality assessment employs mathematical and computational theory to objectively assess the quality of output images based on the human visual system (HVS). In this paper, a novel approach based on multifeature extraction in the spatial and frequency domains is proposed. We combine th...
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/150418 |
| _version_ | 1824455116001902592 |
|---|---|
| author | Tang, Zhisen Zheng, Yuanlin Gu, Ke Liao, Kaiyang Wang, Wei Yu, Miaomiao |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Tang, Zhisen Zheng, Yuanlin Gu, Ke Liao, Kaiyang Wang, Wei Yu, Miaomiao |
| author_sort | Tang, Zhisen |
| collection | NTU |
| description | Objective image quality assessment employs mathematical and computational theory to objectively assess the quality of output images based on the human visual system (HVS). In this paper, a novel approach based on multifeature extraction in the spatial and frequency domains is proposed. We combine the gradient magnitude and phase congruency maps to generate a local structure (LS) map, which can perceive local structural distortions. The LS matches well with HVS and highlights differences with details. For complex visual information, such as texture and contrast sensitivity, we deploy the log-Gabor filter, and spatial frequency, respectively, to effectively capture their variations. Moreover, we employ the random forest (RF) to overcome the limitations of existing pooling methods. Compared with support vector regression, RF can obtain better prediction results. Extensive experimental results on the five benchmark databases indicate that the proposed method precedes all the state-of-the-art image quality assessment metrics in terms of prediction accuracy. In addition, the proposed method is in compliance with the subjective evaluations. |
| first_indexed | 2025-02-19T03:33:05Z |
| format | Journal Article |
| id | ntu-10356/150418 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2025-02-19T03:33:05Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | ntu-10356/1504182021-05-31T01:09:47Z Full-reference image quality assessment by combining features in spatial and frequency domains Tang, Zhisen Zheng, Yuanlin Gu, Ke Liao, Kaiyang Wang, Wei Yu, Miaomiao School of Computer Science and Engineering Engineering::Computer science and engineering Image Quality Assessment (IQA) Log-Gabor Filter Objective image quality assessment employs mathematical and computational theory to objectively assess the quality of output images based on the human visual system (HVS). In this paper, a novel approach based on multifeature extraction in the spatial and frequency domains is proposed. We combine the gradient magnitude and phase congruency maps to generate a local structure (LS) map, which can perceive local structural distortions. The LS matches well with HVS and highlights differences with details. For complex visual information, such as texture and contrast sensitivity, we deploy the log-Gabor filter, and spatial frequency, respectively, to effectively capture their variations. Moreover, we employ the random forest (RF) to overcome the limitations of existing pooling methods. Compared with support vector regression, RF can obtain better prediction results. Extensive experimental results on the five benchmark databases indicate that the proposed method precedes all the state-of-the-art image quality assessment metrics in terms of prediction accuracy. In addition, the proposed method is in compliance with the subjective evaluations. 2021-05-31T01:09:47Z 2021-05-31T01:09:47Z 2019 Journal Article Tang, Z., Zheng, Y., Gu, K., Liao, K., Wang, W. & Yu, M. (2019). Full-reference image quality assessment by combining features in spatial and frequency domains. IEEE Transactions On Broadcasting, 65(1), 138-151. https://dx.doi.org/10.1109/TBC.2018.2871376 0018-9316 0000-0001-9888-928X 0000-0001-5540-3235 0000-0002-6175-4602 https://hdl.handle.net/10356/150418 10.1109/TBC.2018.2871376 2-s2.0-85054496419 1 65 138 151 en IEEE Transactions on Broadcasting © 2018 IEEE. All rights reserved. |
| spellingShingle | Engineering::Computer science and engineering Image Quality Assessment (IQA) Log-Gabor Filter Tang, Zhisen Zheng, Yuanlin Gu, Ke Liao, Kaiyang Wang, Wei Yu, Miaomiao Full-reference image quality assessment by combining features in spatial and frequency domains |
| title | Full-reference image quality assessment by combining features in spatial and frequency domains |
| title_full | Full-reference image quality assessment by combining features in spatial and frequency domains |
| title_fullStr | Full-reference image quality assessment by combining features in spatial and frequency domains |
| title_full_unstemmed | Full-reference image quality assessment by combining features in spatial and frequency domains |
| title_short | Full-reference image quality assessment by combining features in spatial and frequency domains |
| title_sort | full reference image quality assessment by combining features in spatial and frequency domains |
| topic | Engineering::Computer science and engineering Image Quality Assessment (IQA) Log-Gabor Filter |
| url | https://hdl.handle.net/10356/150418 |
| work_keys_str_mv | AT tangzhisen fullreferenceimagequalityassessmentbycombiningfeaturesinspatialandfrequencydomains AT zhengyuanlin fullreferenceimagequalityassessmentbycombiningfeaturesinspatialandfrequencydomains AT guke fullreferenceimagequalityassessmentbycombiningfeaturesinspatialandfrequencydomains AT liaokaiyang fullreferenceimagequalityassessmentbycombiningfeaturesinspatialandfrequencydomains AT wangwei fullreferenceimagequalityassessmentbycombiningfeaturesinspatialandfrequencydomains AT yumiaomiao fullreferenceimagequalityassessmentbycombiningfeaturesinspatialandfrequencydomains |