Point spread function for diffuser cameras based on wave propagation and projection model

Area of Science:

• Optics and Photonics
• Computational Imaging

Background:

• Diffuser cameras offer a compact alternative to traditional imaging systems for light field capture.
• Object reconstruction relies on deconvolving sensor images with depth-specific point-spread-functions (PSFs).
• Current PSF acquisition methods are complex and unreliable, hindering diffuser camera applications.

Purpose of the Study:

• To develop a theoretical PSF model for diffuser cameras.
• To enable accurate object reconstruction at any depth.
• To simplify and improve the reliability of diffuser camera imaging.

Main Methods:

• Estimated diffuser phase using a projection model.
• Derived image response at any depth via forward Fourier optics.
• Validated the theoretical PSF model against experimental data.

Main Results:

• High correlation observed between captured and model-derived PSFs.
• Accurate reconstruction of objects at various depths demonstrated.
• Model-derived PSFs achieved reconstruction quality comparable to real captured PSFs.

Conclusions:

• The proposed theoretical PSF model effectively characterizes diffuser camera behavior.
• This model significantly reduces complexity and enhances reliability for diffuser camera imaging.
• Enables robust object reconstruction across different depths, broadening application potential.