.png){kind=icon}
.png){kind=icon}
Return
A Periodic Error Correction Method for Terahertz Coded-Aperture Imaging
.png){kind=icon}
OA
AI
DOI：10.1002/aisy.202401099.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Terahertz coded-aperture imaging technology (TCAI) has attracted considerable attention due to its high-resolution characteristics in the field of staring imaging. Recent research in metasurface have demonstrated its potential for enhancing terahertz applications by enabling precise control over electromagnetic waves. However, traditional methods for image reconstruction often depend on theoretical calculation data, which are susceptible to environmental noise and model inaccuracies. To solve this problem, this article proposes an unsupervised learning method that iteratively corrects these errors through periodic training. This method models the physical process as a constraint in the training work, starting with theoretical calculations of the radiation field matrix to predict scattering coefficients, which reduces data requirements and enhances robustness. The simulation results demonstrate its superior image quality and noise resistance compared to traditional algorithms, especially under low sampling rates, which provides a strong foundation for practical implementations of TCAI.
Keywords:
metasurface
terahertz coded-aperture imaging
unsupervised learning
Journal
IF:
6.1
Papers:
2.0K
Citations:
8.4K