The SSIM index is not a metric and it is badly evaluate the similarity of images
The article explored some properties of a very popular feature of image structural similarity, called the SSIM index. According to https://scholar.google.com, the article [3], where it was first described, has made more than 20,800 citations during the last 14 years. This indicator is actively used...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Belarusian National Technical University
2019-08-01
|
| Series: | Системный анализ и прикладная информатика |
| Subjects: | |
| Online Access: | https://sapi.bntu.by/jour/article/view/261 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832557358847361024 |
|---|---|
| author | V. V. Starovoitov |
| author_facet | V. V. Starovoitov |
| author_sort | V. V. Starovoitov |
| collection | DOAJ |
| description | The article explored some properties of a very popular feature of image structural similarity, called the SSIM index. According to https://scholar.google.com, the article [3], where it was first described, has made more than 20,800 citations during the last 14 years. This indicator is actively used by the scientific community in imaging research. It acquired the status of an unofficial international standard for assessing image quality in the presence of a template, often referred to as the image quality metric. This article debunks some of the myths that have arisen around this index. A theorem is proved which states that the SSIM index and any of its linear transformations are not metric functions. In many publications and in the Matlab application software package in the description of the SSIM function, it is said that the SSIM index is used to measure the image quality. However, this index, as well as any comparison function with a reference image (such as full-reference), in principle, cannot assess the quality of the analyzed images. They estimate only a certain degree of similarity between the template image and its distorted copy. The article also shows that the SSIM index cannot always correctly determine the similarity of images of the same scene, while the Pearson linear correlation coefficient makes it much faster and more accurate. |
| format | Article |
| id | doaj-art-2ddcbd085a7b43d891765d8b4b4ac3d2 |
| institution | Kabale University |
| issn | 2309-4923 2414-0481 |
| language | English |
| publishDate | 2019-08-01 |
| publisher | Belarusian National Technical University |
| record_format | Article |
| series | Системный анализ и прикладная информатика |
| spelling | doaj-art-2ddcbd085a7b43d891765d8b4b4ac3d22025-02-03T05:16:51ZengBelarusian National Technical UniversityСистемный анализ и прикладная информатика2309-49232414-04812019-08-0102121710.21122/2309-4923-2019-2-12-17199The SSIM index is not a metric and it is badly evaluate the similarity of imagesV. V. Starovoitov0United Institute of Informatics Problems, National Academy of Sciences of BelarusThe article explored some properties of a very popular feature of image structural similarity, called the SSIM index. According to https://scholar.google.com, the article [3], where it was first described, has made more than 20,800 citations during the last 14 years. This indicator is actively used by the scientific community in imaging research. It acquired the status of an unofficial international standard for assessing image quality in the presence of a template, often referred to as the image quality metric. This article debunks some of the myths that have arisen around this index. A theorem is proved which states that the SSIM index and any of its linear transformations are not metric functions. In many publications and in the Matlab application software package in the description of the SSIM function, it is said that the SSIM index is used to measure the image quality. However, this index, as well as any comparison function with a reference image (such as full-reference), in principle, cannot assess the quality of the analyzed images. They estimate only a certain degree of similarity between the template image and its distorted copy. The article also shows that the SSIM index cannot always correctly determine the similarity of images of the same scene, while the Pearson linear correlation coefficient makes it much faster and more accurate.https://sapi.bntu.by/jour/article/view/261image quality assessmentthe ssim indexpearson’s correlation coefficient. |
| spellingShingle | V. V. Starovoitov The SSIM index is not a metric and it is badly evaluate the similarity of images Системный анализ и прикладная информатика image quality assessment the ssim index pearson’s correlation coefficient. |
| title | The SSIM index is not a metric and it is badly evaluate the similarity of images |
| title_full | The SSIM index is not a metric and it is badly evaluate the similarity of images |
| title_fullStr | The SSIM index is not a metric and it is badly evaluate the similarity of images |
| title_full_unstemmed | The SSIM index is not a metric and it is badly evaluate the similarity of images |
| title_short | The SSIM index is not a metric and it is badly evaluate the similarity of images |
| title_sort | ssim index is not a metric and it is badly evaluate the similarity of images |
| topic | image quality assessment the ssim index pearson’s correlation coefficient. |
| url | https://sapi.bntu.by/jour/article/view/261 |
| work_keys_str_mv | AT vvstarovoitov thessimindexisnotametricanditisbadlyevaluatethesimilarityofimages AT vvstarovoitov ssimindexisnotametricanditisbadlyevaluatethesimilarityofimages |