Image Restoration for Tilt Imaging Using Optical Transfer Function Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image restoration methods for tilt imaging suffer from adverse effects like insufficient correction, excessive correction, edge fall, and ringing due to rotationally asymmetrical aberrations, and require a large information amount to correct image quality issues.
Innovation Solution
An image processing method that acquires inclination and aberration information of the imaging optical system, sets a correction amount based on this information, and uses image restoration filters to perform image restoration processing, reducing data requirements by rotating and developing optical transfer functions symmetrically around the image center or optical axis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image restoration processing is performed using optical transfer function for all positions in the screen to correct rotationally asymmetrical aberration in tilt imaging, then correction accuracy is improved, but information amount is increased
Solution Approach 1:
The optical transfer function is segmented into two components: a rotationally symmetrical component (first optical transfer function) and a rotationally asymmetrical component (second optical transfer function). This segmentation allows the asymmetrical aberration to be corrected using only the asymmetrical component, reducing the information amount required while maintaining correction accuracy.
Solution Approach 2:
The rotationally asymmetrical component of the optical transfer function is extracted from the complete optical transfer function. By using only this extracted asymmetrical component for correction, the patent reduces the information amount needed while specifically targeting the rotationally asymmetrical aberration caused by tilt imaging.
2Loss of information
If image restoration processing is performed using rotationally symmetrical optical transfer function on tilt imaging, then information amount is reduced, but adverse effects such as insufficient correction, excessive correction, edge fall, and ringing occur
Solution Approach 1:
The patent applies different optical transfer function components to different aspects of the correction process. The rotationally symmetrical component handles general image restoration, while the rotationally asymmetrical component specifically addresses the asymmetrical aberration. This local quality differentiation ensures accurate correction without requiring the full information set.
Solution Approach 2:
The patent explicitly introduces asymmetry by separating the optical transfer function into symmetrical and asymmetrical components. The asymmetrical component is specifically designed to correct the rotationally asymmetrical aberration inherent in tilt imaging, thereby improving correction quality without increasing overall information requirements.
3Duration of action of moving object
If tilt imaging is performed to expand in-focus range, then depth of field is increased, but rotationally asymmetrical aberration occurs deteriorating image forming performance
Solution Approach 1:
The patent converts the harmful rotationally asymmetrical aberration caused by tilt imaging into a beneficial correction target. By extracting and using the asymmetrical component of the optical transfer function, the system specifically compensates for the aberration introduced by tilt imaging, thereby maintaining the expanded in-focus range while restoring image forming performance.
Data Source
AI summary
An image processing method includes acquiring a captured image that has been generated by imaging through an imaging optical system inclined relative to an imaging plane, acquiring inclination information of the imaging optical system, acquiring aberration information of the imaging optical system, and performing image restoration processing on the captured image based on the aberration information. The image restoration processing includes a setting of a correction amount of the captured image according to the inclination information.


