Polarizer-free two-pixel polarimetric camera by compressive sensing.

Area of Science:

• Optics and Photonics
• Computational Imaging
• Signal Processing

Background:

• Polarimetric imaging traditionally requires polarizers, adding complexity and cost.
• Compressive sensing (CS) offers efficient data acquisition but has not been extensively applied to polarimetry without polarizers.
• Digital Micromirror Devices (DMDs) provide programmable optical control, enabling novel imaging architectures.

Purpose of the Study:

• To propose and validate a novel compressive sensing (CS) polarimetric imaging system that eliminates the need for traditional polarizers.
• To develop and evaluate advanced reconstruction algorithms for accurate polarimetric data retrieval from CS measurements.
• To assess the system's performance, including compression rates and reconstruction quality, and analyze the impact of experimental imperfections.

Main Methods:

• A polarimetric imaging setup utilizing a digital micromirror (DMD) array and two single-pixel detectors was designed.
• A combined reconstruction approach, integrating CS and source separation, was developed and implemented using a reweighted fast iterative shrinkage-thresholding algorithm.
• Physical constraints specific to polarimetric imaging were incorporated into a constrained generalized forward-backward algorithm to further enhance reconstruction.

Main Results:

• Numerical simulations demonstrated the effectiveness of the two-pixel CS polarimetric imaging setup.
• The proposed combined reconstruction method significantly outperformed a two-step approach.
• High compression rates and good reconstruction quality were achieved for polarimetric contrast data.
• The influence of DMD experimental imperfections was analyzed, and 2D polarimetric imaging results were presented.

Conclusions:

• The developed CS polarimetric imaging system offers a polarizer-free and efficient alternative for acquiring polarimetric data.
• Advanced reconstruction algorithms, particularly the combined and constrained methods, are crucial for optimal performance.
• The system shows promise for various applications requiring compact and high-performance polarimetric imaging.