Point spread function for the wide-field-of-view plenoptic cameras

Area of Science:

• Optics
• Image Processing
• Computational Photography

Background:

• Plenoptic cameras with monocentric lenses and microlens arrays (MLAs) are explored for expanding field of view (FOV).
• Accurate modeling of point spread functions (PSFs) is crucial for enhancing imaging quality and light field reconstruction in these systems.
• Existing models do not adequately address the complexities of wide-FOV plenoptic cameras incorporating monocentric optics.

Purpose of the Study:

• To develop a generic image formation model for wide-FOV plenoptic cameras utilizing monocentric lenses and MLAs.
• To derive accurate point spread functions (PSFs) for improved image quality and light field reconstruction.
• To validate the proposed PSF model through experimental comparison.

Main Methods:

• Approximating the monocentric lens by a series of concentric lenses with variable apertures.
• Modeling the equivalent imaging process using geometry simplification and wave propagation.
• Deriving the PSF based on the modeled imaging process.
• Validating the model by comparing generated PSFs with experimentally captured ones.
• Applying image reconstruction through deconvolution using the derived PSFs.

Main Results:

• A novel PSF model for wide-FOV plenoptic cameras with monocentric lenses and MLAs was successfully developed.
• The proposed model accurately represents the optical characteristics and generates valid PSFs.
• Image reconstruction via deconvolution with the proposed PSFs significantly improved image quality compared to subaperture images.

Conclusions:

• The developed PSF model is effective for wide-FOV plenoptic cameras.
• The model facilitates substantial improvements in imaging quality and light field reconstruction.
• This work provides a valuable tool for advancing plenoptic camera technology.