Resolution enhancement for fiber bundle imaging using maximum a posteriori estimation

Area of Science:

• Optics and Imaging
• Computational Imaging
• Image Reconstruction

Background:

• Coherent fiber bundle imaging systems often suffer from limited spatial resolution and fixed structural patterns.
• Existing methods may not effectively address both resolution enhancement and pattern removal simultaneously.

Purpose of the Study:

• To develop a novel framework for joint spatial resolution improvement and fixed pattern removal in coherent fiber bundle imaging.
• To leverage principled forward modeling and maximum a posteriori (MAP) estimation for enhanced image reconstruction.

Main Methods:

• A principled forward model was developed, mapping high-resolution representations to low-resolution images considering probe motions and point spread functions.
• A smoothing prior was incorporated into the forward model.
• Maximum a posteriori (MAP) estimation was computed using conjugate gradient descent to reconstruct high-resolution images from low-resolution inputs.
• The framework supports various applicable transformations.

Main Results:

• The proposed method successfully enhanced spatial resolution and removed fixed structural patterns.
• On USAF target data with 16 low-resolution captures, a spatial resolution enhancement factor of 2.8 was achieved.
• Validation was performed on synthesized data, a transmissive USAF target, and lens tissue data.

Conclusions:

• The developed framework offers a robust approach for improving image quality in coherent fiber bundle imaging systems.
• Principled forward modeling and MAP estimation provide a powerful combination for addressing limitations in this imaging domain.