Global deblurring for continuous out-of-focus images using a depth-varying diffusion model

Area of Science:

• Optical Microscopy
• Image Processing
• Computational Imaging

Background:

• High-magnification microscopy often produces continuous out-of-focus images of large targets.
• Traditional deblurring methods struggle with depth-varying point spread functions (DVPSFs), hindering global analysis.

Purpose of the Study:

• To develop a global deblurring method for continuous out-of-focus images of large-scale spherical samples.
• To enable accurate observation of global optical features in high-magnification microscopy.

Main Methods:

• Analyzed energy diffusion characteristics and developed a 3D continuous energy diffusion model for optical imaging.
• Proposed an adaptive weight depth calculation method considering surface curvature and light direction.
• Developed a universal deblurring method for continuous out-of-focus images of large-scale sphere samples.

Main Results:

• The method accurately calculates sample surface depth and energy diffusion parameters at each depth.
• Successfully achieved image deblurring for continuously changing surfaces.
• Demonstrated global deblurring of multiple samples within a wide field of view using dynamic microspheres.

Conclusions:

• The proposed method effectively addresses challenges in deblurring continuous out-of-focus microscopy images.
• Enables comprehensive global observation of optical features in large-scale samples.
• Validates effectiveness using dynamic microspheres of varying sizes.