Designing Illumination Patterns for Single-Pixel Imaging Using Lattice Models

TL;DR

This paper introduces a novel approach to designing illumination patterns for single-pixel imaging by applying lattice models from statistical mechanics, aiming to improve efficiency and adaptability across various imaging modalities.

Contribution

It establishes a systematic method using spin lattice models to generate structured illumination patterns, bridging statistical mechanics and imaging technology.

Findings

Lattice models can produce effective illumination patterns for single-pixel imaging.

The approach is adaptable for binary, grayscale, and color imaging.

Systematic pattern design can enhance imaging efficiency.

Abstract

Single-pixel imaging leverages a single-pixel detector and structured illumination patterns to reconstruct images, offering a cost-effective solution for imaging across a wide range of wavelengths, such as x-ray and terahertz. However, the technique faces challenges in efficiency due to the need for numerous patterns to achieve high-quality image reconstruction. In this study, we explore the use of spin lattice models from statistical mechanics to design illumination patterns for single-pixel imaging. By employing models like Ising, Potts, XY, and Heisenberg, we generate structured patterns that are adaptable for binary, grayscale, and color imaging. This work creates a direct connection between lattice models and imaging applications, providing a systematic approach to pattern generation that can enhance single-pixel imaging efficiency.