Iterative Decoupling Method for High-Precision Imaging of Complex Surfaces
Nonlinear systems and interaction forces are pervasive in many scientific fields, such as nanoscale metrology and materials science, but their accurate identification is challenging due to their complex behaviour and inaccessibility of measured domains. This problem intensifies for continuous systems undergoing distributed, coupled interactions, such as in the case of topography measurement systems, measuring narrow and deep grooves. Presented is a method to invert a set of nonlinear coupled equations, which can be functions of unknown distributed physical quantities. The method employs a successive approach to iteratively converge to the exact solution of the set of nonlinear equations. The latter utilizes an approximate yet invertible model providing an inexact solution, which is evaluated using the hard-to-invert exact model of the system. This method is applied to the problem of reconstructing the topography of surface contours using a thin and long vibrating fiber. In nanoscale metrology, measuring inaccessible deep and narrow grooves or steep walls becomes difficult and singular when attempting to extract distributed nonlinear interactions that depend on the topography. We verify our method numerically by simulating the Van der Waals (VdW) interaction forces between a nanofiber and a nanoscale deep groove, and experimentally by exploiting magnetic interactions between a magnetic topography and a vibrating, elastic beam. Our results validate the ability to accurately reconstruct the topography of normally inaccessible regions, making it a possible enhancement for traditional point based AFM measurements, as well as for other nonlinear inverse problems.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Bridging the Arithmetic Gap: The Cognitive Complexity Benchmark and Financial-PoT for Robust Financial Reasoning
While Large Language Models excel at semantic tasks, they face a critical bottleneck in financial quantitative reasoning, frequently suffering from "Arithmetic Hallucinations" and a systemic failure mode we term "Cogniti…
The R2D2 deep neural network series paradigm for fast precision imaging in radio astronomy
Radio-interferometric (RI) imaging entails solving high-resolution high-dynamic range inverse problems from large data volumes. Recent image reconstruction techniques grounded in optimization theory have demonstrated rem…
AstronomyImage ReconstructionCLEANing Cygnus A deep and fast with R2D2
A novel deep learning paradigm for synthesis imaging by radio interferometry in astronomy was recently proposed, dubbed "Residual-to-Residual DNN series for high-Dynamic range imaging" (R2D2). In this work, we start by s…
AstronomyComputational EfficiencyRadio InterferometryPost-training Quantization for Hybrid Iterative Generative Models
Iterative Generative Models (IGMs) span autoregressive and diffusion paradigms, and hybrid variants that couple them can achieve remarkable image-generation fidelity. However, their iterative inference incurs substantial…
Spectral decoupling allows training transferable neural networks in medical imaging
Many current neural networks for medical imaging generalise poorly to data unseen during training. Such behaviour can be caused by networks overfitting easy-to-learn, or statistically dominant, features while disregardin…
Image Classificationwhole slide images