Computational ghost imaging with low-density parity-check code.

Area of Science:

• Computational imaging
• Optical sensing technologies
• Information theory in signal processing

Background:

• Ghost imaging (GI) provides a robust non-local sensing capability across diverse wavebands.
• Existing GI research primarily addresses reconstruction from undersampled signals.
• Exploiting information redundancy for enhanced reconstruction reliability in complex environments remains underexplored.

Purpose of the Study:

• To introduce a novel computational ghost imaging system utilizing Low-Density Parity-Check (LDPC)-coded radiation fields.
• To investigate the potential of signal redundancy for suppressing environmental noise and non-ideal factors in GI.
• To establish an analytical framework for evaluating the performance of coded GI systems.

Main Methods:

• Development of a computational GI system incorporating LDPC-coded radiation fields.
• Modeling the GI process as a matching fading channel to analyze signal redundancy.
• Derivation of an analytical lower bound for the bit error rate (BER) of the proposed system.
• Validation through numerical simulations and experimental implementations.

Main Results:

• Demonstration that signal redundancy within the LDPC-coded GI framework effectively suppresses environmental noise and non-ideal factors.
• Establishment of an analytical lower bound for the BER, providing theoretical performance guarantees.
• Significant performance enhancement observed in challenging sensing conditions compared to conventional GI methods.

Conclusions:

• The proposed LDPC-coded GI framework offers a powerful approach to enhance reconstruction reliability in complex sensing environments.
• Information redundancy, when strategically employed via LDPC coding, is crucial for robust ghost imaging.
• The findings pave the way for more resilient and reliable non-local sensing applications.