An Improved SRGAN-Based Deblurring Model for Multiple Blurriness in Microscopy
- 30 September 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-13
Abstract
In microscopy, eliminating blurriness and obtaining a sharp and accurate edge of observed objects in multiple blurriness is a key issue. Herein, a deblurring model called De-SRGAN is proposed, in which multiple discriminators are set. The multi-discriminators architecture can provide stable training, which helps achieve effective deblurring results in multiple blurriness. Moreover, considering the property of the edge in the frequency domain, a perceptual loss term in the frequency domain is constructed as additional factor. Results show that De-SRGAN can generate deblurred images with low MSE/ MAE, high PSNR/SSIM, high perceptual quality, and high accuracy. This study contributes to improving the accuracy of size measurement due to the deblurred images with sharp and accurate edges.Keywords
Funding Information
- National Natural Science Foundation of China (51975102)
- Liaoning Revitalization Talents Program (XLYC2002020)
- Defense Industrial Technology Development Program (JCKY2022203B006)
This publication has 32 references indexed in Scilit:
- Deblurring microscopic image by integrated convolutional neural networkPrecision Engineering, 2023
- Fast MRI reconstruction using StrainNet with dual-domain loss on spatial and frequency spacesIntelligent Systems with Applications, 2023
- Real-aware motion deblurring using multi-attention CycleGAN with contrastive guidanceDigital Signal Processing, 2023
- Joint strong edge and multi-stream adaptive fusion network for non-uniform image deblurringJournal of Visual Communication and Image Representation, 2022
- DPNet: Detail-preserving image deraining via learning frequency domain knowledgeDigital Signal Processing, 2022
- A survey of deep learning approaches to image restorationNeurocomputing, 2022
- Generative Adversarial Networks (GANs) in networking: A comprehensive survey & evaluationComputer Networks, 2021
- Single-image deblurring with neural networks: A comparative surveyComputer Vision and Image Understanding, 2021
- Numerical analysis on high resolution optical measurement method with long working distance objective for in-line inspection of micro-structured surfacePrecision Engineering, 2021
- Development of high-accuracy edge line estimation algorithms using terrestrial laser scanningAutomation in Construction, 2019