An Automatic and Optimal MPA Design Method.

Area of Science:

• Optics and Photonics
• Image Processing
• Materials Science

Background:

• Micro-polarizer arrays (MPAs) are critical components in focal plane polarimetric imaging systems.
• Current MPA design relies heavily on expert experience and limited optimization, often focusing solely on maximizing bandwidth.
• Existing design processes involve exhaustive searches over constrained sets of polarizing patterns, limiting potential improvements.

Purpose of the Study:

• To develop a fully automated and optimal MPA design method (AO-MPA).
• To move beyond single-criterion optimization by incorporating six design principles into a multi-objective optimization framework.
• To significantly enhance MPA performance and manufacturability.

Main Methods:

• Formulation of a tri-objective optimization problem with multiple constraints for MPA design.
• Application of advanced multi-objective optimization techniques for rapid and automated searching of a larger pattern space.
• Comparative experimental testing of AO-MPA against existing design methods and state-of-the-art patterns using a public benchmark dataset.

Main Results:

• AO-MPA demonstrates high computational efficiency and identifies all optimal MPA patterns for various array sizes.
• The method simultaneously obtains all optimal layouts for each size.
• Generated MPAs require fewer polarization orientations and offer superior performance in estimating intensity, Stokes vector, and degree of linear polarization.

Conclusions:

• The AO-MPA method provides a superior, automated approach to MPA design compared to traditional methods.
• The resulting MPAs are easier to manufacture due to fewer required orientations and exhibit enhanced robustness against noise.
• This work advances the field of polarimetric imaging through optimized MPA design.