https://github.com/amatofrancesco99/image-compression-and-subjective-quality-assessment

The aim is to apply a degradation algorithm on 10 images (JPEG compression), ask some users their personal opinion on the quality of that impaired images and perform some statistics on obtained data.
https://github.com/amatofrancesco99/image-compression-and-subjective-quality-assessment

compression-implementations image jpeg-image-compression quality-assessment

Last synced: 7 months ago
JSON representation

The aim is to apply a degradation algorithm on 10 images (JPEG compression), ask some users their personal opinion on the quality of that impaired images and perform some statistics on obtained data.

• Host: GitHub
• URL: https://github.com/amatofrancesco99/image-compression-and-subjective-quality-assessment
• Owner: Amatofrancesco99
• License: mit
• Created: 2021-11-22T12:54:26.000Z (almost 4 years ago)
• Default Branch: main
• Last Pushed: 2021-11-22T14:35:07.000Z (almost 4 years ago)
• Last Synced: 2025-01-23T03:44:36.553Z (9 months ago)
• Topics: compression-implementations, image, jpeg-image-compression, quality-assessment
• Language: Python
• Homepage: https://www.amatofrancesco.altervista.org
• Size: 5.66 MB
• Stars: 0
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Image-compression-and-subjective-quality-assessment

[![forthebadge made-with-python](http://ForTheBadge.com/images/badges/made-with-python.svg)](https://www.python.org/)




[![PyPI license](https://img.shields.io/pypi/l/ansicolortags.svg)](https://github.com/Amatofrancesco99/Image-compression-and-subjective-quality-assessment/blob/main/LICENSE)







The aim of this project is to apply a [two-level degradation algorithm](https://github.com/Amatofrancesco99/Image-compression-and-subjective-quality-assessment/tree/main/main.py) on [10 images](https://github.com/Amatofrancesco99/Image-compression-and-subjective-quality-assessment/tree/main/Original%20images) ([JPEG compression](https://www.ece.ucdavis.edu/cerl/reliablejpeg/compression/)).




After applying this algorithm, some users (24) were asked to give their personal opinion (opinion score, OS) on the quality of such damaged images ([module link](https://docs.google.com/forms/d/1sbGq9buDY81KzAQer61JFAOOB_Cx7w-6KJtURkBYlQM/edit?usp=sharing)).




Subsequently, some [statistics](https://docs.google.com/spreadsheets/d/1KzE92K5fdU3O7zQb6gbgB_71cHL2H_fhqHaPDHUbvUY/edit?usp=sharing) were automatically made on these obtained data.




Finally, the results are studied in depth, to better understand why some images have confused most users.





**Some statistic results**




This table shows for each image its *[MOS](https://www.twilio.com/docs/glossary/what-is-mean-opinion-score-mos)* (μ, [mean](https://en.wikipedia.org/wiki/Mean) opinion score) and *[variance](https://en.wikipedia.org/wiki/Variance)* (σ2).




Moreover, for each impairment group (original, imp1, imp2) is performed its overall MOS and variance.









This graph shows the same results discussed in the previous table, in a different way (through a multi-line graph).









We can see that none of this images has a strange behaviour in its MOS




(it means MOSimp2 < MOSimp1 < MOSoriginal).






**Presentation link**




[![Google](https://img.shields.io/badge/google-4285F4?style=for-the-badge&logo=google&logoColor=white)](https://docs.google.com/presentation/d/13nk13bs2LWd0VW3bJVGlTVi19dvzVleI0sHtMW1I2rw/edit?usp=sharing)