ADMM-inspired image reconstruction for terahertz off-axis digital holography.

Area of Science:

• Optics and Photonics
• Computational Imaging
• Terahertz Science and Technology

Background:

• Image reconstruction in off-axis terahertz digital holography faces challenges due to harsh recording conditions and problem non-convexity.
• Existing methods may struggle with artifacts arising from boundary condition mismatches.

Purpose of the Study:

• To develop an advanced inverse problem-based technique for terahertz digital holography image reconstruction.
• To jointly reconstruct the object field and reference field amplitude for improved accuracy.

Main Methods:

• Utilizing wavelet domain regularization to promote sparse object solutions.
• Employing an alternating direction method of multipliers (ADMM) framework for optimization.
• Implementing a windowing process to mitigate artifacts from camera frame truncation.

Main Results:

• The proposed method effectively reconstructs terahertz digital holography images.
• Demonstrated improvements in reconstruction quality compared to existing techniques.
• Successful joint reconstruction of object and reference fields.

Conclusions:

• The inverse problem-based approach with wavelet regularization offers a robust solution for terahertz digital holography.
• The ADMM-based optimization framework efficiently handles the complex reconstruction problem.
• The method shows significant potential for enhancing imaging applications in the terahertz spectrum.