Quantitative THz time-domain single-pixel imaging using statistical scale-adaptive computational ghost imaging with...

Area of Science:

• Terahertz spectroscopy
• Single-pixel imaging
• Computational ghost imaging

Background:

• Terahertz time-domain spectroscopy (THz-TDS) integrated with single-pixel imaging (SPI) offers simultaneous spatial and spectral information acquisition.
• Computational ghost imaging (CGI) excels in noise robustness but provides relative, not quantitative, measurements.
• Ordinary least squares (OLS) reconstruction offers quantitative scaling but is sensitive to noise.

Purpose of the Study:

• To develop a novel method overcoming the trade-offs between noise robustness and quantitative accuracy in THz-TDS imaging.
• To enable accurate amplitude and phase mapping for spatially resolved spectroscopic analysis.
• To provide a practical and low-cost solution for quantitative THz spectroscopic imaging.

Main Methods:

• Introduction of a statistical scale-adaptive computational ghost imaging (SSA-CGI) method.
• Hybridization of CGI's correlation-based stability with OLS's quantitative scaling.
• Statistical transfer of quantitative voltage scale from OLS to high-SNR CGI reconstruction.

Main Results:

• Experimental validation using metallic apertures and SU-8 structures.
• Demonstration of spatially resolved spectroscopic imaging with high scaling fidelity.
• Obtained group refractive indices consistent with reported values.

Conclusions:

• The SSA-CGI method successfully combines noise robustness and quantitative accuracy.
• This technique offers a practical, low-cost approach for quantitative THz spectroscopic imaging.
• Potential applications include material characterization, nondestructive testing, and biomedical diagnostics.