Resolution analysis in computational imaging with patterned illumination and bucket detection.

Area of Science:

• Computational imaging
• Optical imaging systems
• Pattern projection microscopy

Background:

• Computational imaging utilizes programmable spatial light modulators to project microstructured patterns for object sampling.
• These patterns serve as generalized measurement modes to extract object information.

Purpose of the Study:

• To demonstrate that image resolution in pattern projection computational imaging is solely limited by the numerical aperture of the projection optics.
• To show the superiority of a single-pixel detection strategy over conventional array detectors, especially under adverse optical conditions.

Main Methods:

• Employing a sequence of microstructured light patterns projected via a spatial light modulator.
• Utilizing a single-pixel detection strategy for data acquisition.
• Conducting proof-of-principle experiments to validate theoretical findings.

Main Results:

• Achieved image resolution limited only by the numerical aperture of the projection optics, irrespective of collection optics quality.
• Demonstrated that single-pixel detection outperforms conventional optical array detectors in resolution.
• Successfully retrieved images even with an optical diffuser that hindered conventional megapixel cameras.

Conclusions:

• The numerical aperture of projection optics is the sole determinant of resolution in this computational imaging approach.
• Single-pixel detection offers a robust and high-resolution alternative to array detectors, particularly in challenging optical environments.
• This method enables effective image retrieval under conditions where traditional imaging fails.