Depth Reconstruction for Reference-Free THz Holography Based on Physics-Informed Deep Learning
- Resource Type
- Conference
- Authors
- Xiang, Mingjun; Yuan, Hui; Wang, Lingxiao; Zhou, Kai; Roskos, Hartmut G.
- Source
- 2023 48th International Conference on Infrared, Millimeter, and Terahertz Waves (IRMMW-THz) Infrared, Millimeter, and Terahertz Waves (IRMMW-THz), 2023 48th International Conference on. :1-2 Sep, 2023
- Subject
- Aerospace
Bioengineering
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Engineered Materials, Dielectrics and Plasmas
Fields, Waves and Electromagnetics
Photonics and Electrooptics
Signal Processing and Analysis
Deep learning
Knowledge engineering
Three-dimensional displays
Diffraction
Artificial neural networks
Holography
Reconstruction algorithms
- Language
- ISSN
- 2162-2035
This paper demonstrates a depth reconstruction method for THz holography based on deep learning (DL) algorithm. We incorporate Fresnel diffraction as the physical prior knowledge to obtain a sequence of different intensities with the corresponding depths to train the neural network (NN). The depth reconstruction tasks are achieved based on the dataset transplanted from MNIST. With this approach, we avoid the prohibitively time-consuming collection of a large number of THz-frequency images. Both simulated and experimental results illustrate the accuracy of the method, representing the first steps towards fast THz 3D imaging with reference-beam-free low-cost power detection.