3D image reconstruction of terahertz computed tomography at sparse angles by total variation minimization

Area of Science:

• Physics
• Imaging Science
• Electrical Engineering

Background:

• Continuous-wave terahertz computed tomography (THz-CT) is a key 3D imaging technique.
• Probing sample structure and inner properties requires high-fidelity imaging.
• Data acquisition time is a significant limitation in THz-CT.

Purpose of the Study:

• To reduce data acquisition time in THz-CT by 75%.
• To maintain reconstruction fidelity despite data reduction.
• To validate the effectiveness of the total variation (TV) minimization algorithm for sparse angle reconstruction.

Main Methods:

• Implementation of a THz-CT system utilizing a 278.6 GHz avalanche diode source.
• Generation of a propagation-invariant zero-order Bessel beam using an axicon.
• Application of the total variation (TV) minimization iterative algorithm for image reconstruction.

Main Results:

• Achieved a 75% reduction in THz-CT data acquisition time.
• Demonstrated high reconstruction fidelity comparable to full data acquisition.
• Successfully suppressed noise, artifacts, and shape distortion in reconstructed images.

Conclusions:

• The TV-minimization algorithm significantly accelerates THz-CT data acquisition.
• This method enables efficient 3D imaging of optically opaque objects without compromising image quality.
• Sparse angle reconstruction with TV minimization is a viable approach for THz-CT.