Technology Category
Computer Vision: Other Foundations of Computer VisionMachine Learning: Structured LearningSearch and Optimization: Non-convex Optimization
A Fourier-Based Global Denoising Model for Smart Artifacts Removing of Microscopy Images
🤖 AI Summary
Microscopy images (STM/AFM/SEM) are commonly degraded by noise and artifacts, causing physically significant yet weak microstructural features to be erroneously discarded as noise. Conventional denoising methods rely excessively on intensity-based criteria, failing to simultaneously suppress artifacts and preserve faint signals. To address this, we propose a Fourier-transform-based global denoising model featuring a non-paired dual-channel input scheme that integrates a U-Net architecture with fast Fourier transform (FFT). A pixel-frequency hybrid loss function is designed, incorporating domain-specific priors to explicitly guide weak-signal preservation. Additionally, a preprocessing trade-off strategy enhances robustness against diverse degradation patterns. Extensive evaluation across multiple microscopy modalities demonstrates that our method intelligently suppresses artifacts while faithfully retaining sub-pixel-scale physical structures, thereby significantly improving the accuracy of quantitative microstructural analysis.
Technology Category
Application Category
📝 Abstract
Microscopy such as Scanning Tunneling Microscopy (STM), Atomic Force Microscopy (AFM) and Scanning Electron Microscopy (SEM) are essential tools in material imaging at micro- and nanoscale resolutions to extract physical knowledge and materials structure-property relationships. However, tuning microscopy controls (e.g. scanning speed, current setpoint, tip bias etc.) to obtain a high-quality of images is a non-trivial and time-consuming effort. On the other hand, with sub-standard images, the key features are not accurately discovered due to noise and artifacts, leading to erroneous analysis. Existing denoising models mostly build on generalizing the weak signals as noises while the strong signals are enhanced as key features, which is not always the case in microscopy images, thus can completely erase a significant amount of hidden physical information. To address these limitations, we propose a global denoising model (GDM) to smartly remove artifacts of microscopy images while preserving weaker but physically important features. The proposed model is developed based on 1) first designing a two-imaging input channel of non-pair and goal specific pre-processed images with user-defined trade-off information between two channels and 2) then integrating a loss function of pixel- and fast Fourier-transformed (FFT) based on training the U-net model. We compared the proposed GDM with the non-FFT denoising model over STM-generated images of Copper(Cu) and Silicon(Si) materials, AFM-generated Pantoea sp.YR343 bio-film images and SEM-generated plastic degradation images. We believe this proposed workflow can be extended to improve other microscopy image quality and will benefit the experimentalists with the proposed design flexibility to smartly tune via domain-experts preferences.
Problem
Research questions and friction points this paper is trying to address.
Removing noise and artifacts from microscopy images
Preserving weak but physically important features
Improving image quality without manual parameter tuning
Innovation
Methods, ideas, or system contributions that make the work stand out.
Global denoising model using Fourier transforms
Two-channel input with user-defined trade-offs
U-net trained with pixel and FFT loss
🔎 Similar Papers
No similar papers found.
