Object reconstruction from photon-limited centroided data of randomly translating images

Area of Science:

• Optics and image processing
• Computational imaging

Background:

• Low light level imaging presents challenges for object reconstruction.
• Random image translation complicates traditional reconstruction techniques.

Purpose of the Study:

• To investigate centroiding as a computationally fast object reconstruction technique.
• To develop a method for phase retrieval in low light conditions.

Main Methods:

• Utilizing centroiding on average N-photon frames for image reconstruction.
• Deriving a relationship between the spectrum of centroided frames and the object spectrum.
• Developing and testing a phase retrieval algorithm for 1D objects using computer-simulated data.

Main Results:

• A direct relationship between the average N-photon centroided frame spectrum and the object spectrum was established.
• An algorithm for retrieving phase information from 1D objects was successfully developed and tested.
• Computer simulations validated the proposed reconstruction relationship and algorithm.

Conclusions:

• Centroiding is a viable and efficient technique for object reconstruction from randomly translating images, especially at low light levels.
• The developed method allows for the recovery of both spectral and phase information of the object.
• This approach offers a promising solution for low light imaging challenges.