Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Survey on Image Quality Assessment: Insights, Analysis, and Future Outlook

Chengqian Ma, Zhengyi Shi, Zhiqiang Lu, Shenghao Xie, Fei Chao, Yao Sui

TL;DR

This survey comprehensively catalogs IQA methods across general and specific scenarios, organizing approaches by SIQA/OIQA and FR/RR/NR categories. It covers traditional statistics, NSS, and ML methods as well as CNN- and Transformer-based models, and discusses training frameworks to address data scarcity. The analysis highlights that while DL-based IQA achieves strong performance, classic metrics like PSNR and SSIM remain popular for their simplicity and interpretability, and there is a need for distortion-aware, domain-specific metrics. The authors argue for practical IQA development that aligns with real-world application requirements, emphasizing interpretability, user-centric evaluation, and targeted improvements for medical, dehazing, portrait, and deburring scenarios.

Abstract

Image quality assessment (IQA) represents a pivotal challenge in image-focused technologies, significantly influencing the advancement trajectory of image processing and computer vision. Recently, IQA has witnessed a notable surge in innovative research efforts, driven by the emergence of novel architectural paradigms and sophisticated computational techniques. This survey delivers an extensive analysis of contemporary IQA methodologies, organized according to their application scenarios, serving as a beneficial reference for both beginners and experienced researchers. We analyze the advantages and limitations of current approaches and suggest potential future research pathways. The survey encompasses both general and specific IQA methodologies, including conventional statistical measures, machine learning techniques, and cutting-edge deep learning models such as convolutional neural networks (CNNs) and Transformer models. The analysis within this survey highlights the necessity for distortion-specific IQA methods tailored to various application scenarios, emphasizing the significance of practicality, interpretability, and ease of implementation in future developments.

A Survey on Image Quality Assessment: Insights, Analysis, and Future Outlook

TL;DR

This survey comprehensively catalogs IQA methods across general and specific scenarios, organizing approaches by SIQA/OIQA and FR/RR/NR categories. It covers traditional statistics, NSS, and ML methods as well as CNN- and Transformer-based models, and discusses training frameworks to address data scarcity. The analysis highlights that while DL-based IQA achieves strong performance, classic metrics like PSNR and SSIM remain popular for their simplicity and interpretability, and there is a need for distortion-aware, domain-specific metrics. The authors argue for practical IQA development that aligns with real-world application requirements, emphasizing interpretability, user-centric evaluation, and targeted improvements for medical, dehazing, portrait, and deburring scenarios.

Abstract

Image quality assessment (IQA) represents a pivotal challenge in image-focused technologies, significantly influencing the advancement trajectory of image processing and computer vision. Recently, IQA has witnessed a notable surge in innovative research efforts, driven by the emergence of novel architectural paradigms and sophisticated computational techniques. This survey delivers an extensive analysis of contemporary IQA methodologies, organized according to their application scenarios, serving as a beneficial reference for both beginners and experienced researchers. We analyze the advantages and limitations of current approaches and suggest potential future research pathways. The survey encompasses both general and specific IQA methodologies, including conventional statistical measures, machine learning techniques, and cutting-edge deep learning models such as convolutional neural networks (CNNs) and Transformer models. The analysis within this survey highlights the necessity for distortion-specific IQA methods tailored to various application scenarios, emphasizing the significance of practicality, interpretability, and ease of implementation in future developments.

Paper Structure

This paper contains 20 sections, 3 figures, 9 tables.

Table of Contents

  1. Introduction
  2. Image Quality Assessment Methods
  3. General Scene Methods
  4. Statistics Methods
  5. HVS-Based Methods
  6. Transform Domain-Based Methods
  7. NSS-Based Methods
  8. Machine Learning-Based Methods
  9. Model-based Methods
  10. Traditional Machine Learning Methods
  11. CNN-Based Methods
  12. Transformer-Based Methods
  13. Framework-Based Methods
  14. Specific Scene Methods
  15. Medical IQA
  16. ...and 5 more sections

Figures (3)

  • Figure 1: Classification of IQA Methods.
  • Figure 2: Publication Times of HVS-based Methods, Transform Domain-based Methods, NSS-based Methods and Traditional Machine Learning Methods. For methods that span more than one line, we enclose them in [].
  • Figure 3: Publication Times of CNN-Based Methods, Transformer-Based Methods, Framework-Based Methods, Specific Scene Methods. For methods that span more than one line, we enclose them in [].