Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Super-Resolution for Interferometric Imaging: Model Comparisons and Performance Analysis

Hasan Berkay Abdioglu, Rana Gursoy, Yagmur Isik, Ibrahim Cem Balci, Taha Unal, Kerem Bayer, Mustafa Ismail Inal, Nehir Serin, Muhammed Furkan Kosar, Gokhan Bora Esmer, Huseyin Uvet

TL;DR

This work investigates overcoming diffraction-imposed resolution limits in holographic microscopy by applying Super-Resolution to interferometric imaging. It compares RCAN (CNN-based) and Real-ESRGAN (GAN-based) on a microparticle holographic dataset, using both image-quality and phase-map metrics. RCAN achieves higher numerical fidelity in phase reconstruction, while Real-ESRGAN delivers superior visual quality, revealing a trade-off between quantitative accuracy and perceptual realism. The study highlights the potential of phase-aware loss functions and application-driven model selection to advance interferometric imaging in biomedical and materials diagnostics.

Abstract

This study investigates the application of Super-Resolution techniques in holographic microscopy to enhance quantitative phase imaging. An off-axis Mach-Zehnder interferometric setup was employed to capture interferograms. The study evaluates two Super-Resolution models, RCAN and Real-ESRGAN, for their effectiveness in reconstructing high-resolution interferograms from a microparticle-based dataset. The models were assessed using two primary approaches: image-based analysis for structural detail enhancement and morphological evaluation for maintaining sample integrity and phase map accuracy. The results demonstrate that RCAN achieves superior numerical precision, making it ideal for applications requiring highly accurate phase map reconstruction, while Real-ESRGAN enhances visual quality and structural coherence, making it suitable for visualization-focused applications. This study highlights the potential of Super-Resolution models in overcoming diffraction-imposed resolution limitations in holographic microscopy, opening the way for improved imaging techniques in biomedical diagnostics, materials science, and other high-precision fields.

Super-Resolution for Interferometric Imaging: Model Comparisons and Performance Analysis

TL;DR

This work investigates overcoming diffraction-imposed resolution limits in holographic microscopy by applying Super-Resolution to interferometric imaging. It compares RCAN (CNN-based) and Real-ESRGAN (GAN-based) on a microparticle holographic dataset, using both image-quality and phase-map metrics. RCAN achieves higher numerical fidelity in phase reconstruction, while Real-ESRGAN delivers superior visual quality, revealing a trade-off between quantitative accuracy and perceptual realism. The study highlights the potential of phase-aware loss functions and application-driven model selection to advance interferometric imaging in biomedical and materials diagnostics.

Abstract

This study investigates the application of Super-Resolution techniques in holographic microscopy to enhance quantitative phase imaging. An off-axis Mach-Zehnder interferometric setup was employed to capture interferograms. The study evaluates two Super-Resolution models, RCAN and Real-ESRGAN, for their effectiveness in reconstructing high-resolution interferograms from a microparticle-based dataset. The models were assessed using two primary approaches: image-based analysis for structural detail enhancement and morphological evaluation for maintaining sample integrity and phase map accuracy. The results demonstrate that RCAN achieves superior numerical precision, making it ideal for applications requiring highly accurate phase map reconstruction, while Real-ESRGAN enhances visual quality and structural coherence, making it suitable for visualization-focused applications. This study highlights the potential of Super-Resolution models in overcoming diffraction-imposed resolution limitations in holographic microscopy, opening the way for improved imaging techniques in biomedical diagnostics, materials science, and other high-precision fields.

Paper Structure

This paper contains 18 sections, 6 equations, 5 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Data Collection System
  4. Data Preprocessing and Degradation Handling
  5. Architecture of Super-Resolution Models
  6. RCAN: Residual Channel Attention Network
  7. Loss Functions for RCAN Training
  8. Real-ESRGAN: Enhanced Super-Resolution Generative Adversarial Network
  9. Loss Functions for Real-ESRGAN Training
  10. Training Optimization Strategies for Improved Performance
  11. Evaluation of Super-Resolution Image Quality
  12. Structural Integrity Metrics
  13. Error Metrics
  14. Results
  15. Evaluation on 2D Image Data
  16. ...and 3 more sections

Figures (5)

  • Figure 1: Schematic Representation of the Mach-Zehnder Interferometer Setup
  • Figure 2: Outputs of Different Downscaling Methodologies
  • Figure 3: Enlarged Visualization of the Super-Resolution Results
  • Figure 4: Comparison of Super-Resolution and Original Images
  • Figure 5: A representative phase profile depicting the optical path difference (OPD) transformation applied to both the original and generated datasets.