Research on super-resolution reconstruction technology of atmospheric turbulence phase screens based on deep learning

Atmospheric turbulence phase screens are commonly utilized to calculate the transmission of light in the atmosphere. The key to accurately generating an atmospheric turbulence phase screen is to satisfy the statistical characteristics of atmospheric turbulence. Traditional methods for generating turbulence phase screens are limited by high computational complexity. Therefore, enhancing the image quality of turbulence phase screens is of utmost importance. This paper presents a super-resolution reconstruction algorithm for atmospheric turbulence phase screens based on deep learning. The proposed model integrates the statistical properties of atmospheric turbulence into the neural network and incorporates the wavefront phase structure function to modify the loss function. Experimental results demonstrate that the atmospheric turbulence phase screens generated by the improved super-resolution convolutional neural networks (ISRCNN) are substantially consistent with theoretical values in terms of statistical characteristics. The proposed method achieves a modest improvement over bicubic interpolation and super-resolution convolutional neural networks (SRCNN) models with respect to statistical properties.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2026-11-16
DOI
https://doi.org/10.1038/s41598-026-73558-9
Primary Topic
Advanced Image Processing Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Research on super-resolution reconstruction technology of atmospheric turbulence phase screens based on deep learning

Mengjie Zhao, Ying Chen
2 links
Scientific Reports
Advanced Image Processing Techniques
article

Research on super-resolution reconstruction technology of atmospheric turbulence phase screens based on deep learning

Mengjie Zhao, Ying Chen
article en

Abstract

Atmospheric turbulence phase screens are commonly utilized to calculate the transmission of light in the atmosphere. The key to accurately generating an atmospheric turbulence phase screen is to satisfy the statistical characteristics of atmospheric turbulence. Traditional methods for generating turbulence phase screens are limited by high computational complexity. Therefore, enhancing the image quality of turbulence phase screens is of utmost importance. This paper presents a super-resolution reconstruction algorithm for atmospheric turbulence phase screens based on deep learning. The proposed model integrates the statistical properties of atmospheric turbulence into the neural network and incorporates the wavefront phase structure function to modify the loss function. Experimental results demonstrate that the atmospheric turbulence phase screens generated by the improved super-resolution convolutional neural networks (ISRCNN) are substantially consistent with theoretical values in terms of statistical characteristics. The proposed method achieves a modest improvement over bicubic interpolation and super-resolution convolutional neural networks (SRCNN) models with respect to statistical properties.

Scientific Reports
Hohai University (CN), Chengdu Technological University (CN), Bohai University (CN)
Advanced Image Processing Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon