Framework for improving DoLP and AoLP reconstruction quality in microgrid polarimeters

Area of Science:

• Optics and Photonics
• Image Processing
• Computational Imaging

Background:

• Polarimetric imaging relies on accurate computation of degree and angle of linear polarization (DoLP and AoLP) from Stokes parameters.
• Snapshot imagers using microgrid polarizer arrays (MPAs) suffer from DoLP and AoLP degradation due to spatial bandwidth mismatches in reconstructed Stokes parameters (S0, S1, S2).

Purpose of the Study:

• To develop a novel demosaicking strategy for microgrid polarizer array (MPA) data that directly estimates DoLP and AoLP.
• To overcome spatial bandwidth mismatch issues inherent in conventional Stokes parameter estimation for snapshot polarimetric imaging.

Main Methods:

• Rigorous analysis showing that log-MPA-sampled data decouples DoLP and AoLP from S0 in the spatial Fourier domain.
• Proposed an alternative demosaicking strategy for direct DoLP and AoLP estimation from MPA data, bypassing Stokes parameter calculation.
• Experimental verification using simulated 2x2 and 2x4 MPA patterns and comparison with five existing demosaicking methods.

Main Results:

• The proposed log-MPA demosaicking method significantly improves DoLP and AoLP reconstruction accuracy compared to conventional Stokes parameter methods.
• The novel approach alleviates spatial bandwidth mismatch problems, leading to reduced computational complexity.
• Closed-form error analysis confirmed negligible approximation errors for practical applications.

Conclusions:

• Log-MPA demosaicking offers a superior approach for accurate DoLP and AoLP estimation in snapshot polarimetric imaging.
• This method effectively addresses limitations of traditional techniques, enhancing image quality and computational efficiency.
• The findings pave the way for more robust and precise polarimetric imaging systems.