Restoration Filter Generation for Image Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image restoration filters, such as the Wiener filter, excel in point image restoration but often degrade image quality by causing ringing and emphasizing return noise, especially in edge areas with rapid contrast changes and in equipment with limited throughput like mobile devices.
Innovation Solution
An image processing device generates a restoration filter by combining multiple base filters that do not cause ringing or emphasize return noise, fitting to an ideal filter based on the point spread function of the optical system, to achieve high accuracy and prevent image quality degradation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional restoration filters (e.g., Wiener filter) are used for point image restoration, then point image restoration accuracy is improved, but image quality excluding point image restoration accuracy deteriorates due to ringing and return noise emphasis
Solution Approach 1:
The restoration filter is segmented into multiple base filters with different characteristics. Each base filter handles specific frequency ranges or image regions, allowing the system to achieve accurate point image restoration while suppressing ringing and return noise through coordinated operation of multiple specialized filters
Solution Approach 2:
The invention changes the parameters of the restoration filter by adjusting the number of taps and coefficients based on image characteristics such as contrast and frequency content. This dynamic parameter adjustment enables the filter to adapt to different image regions, achieving high restoration accuracy while minimizing harmful artifacts like ringing and return noise
2Productivity
If point image restoration processing is performed in mobile devices with limited throughput, then processing speed is improved, but image quality excluding point image restoration accuracy deteriorates due to restricted filter taps
Solution Approach 1:
The restoration filter is made dynamic by adaptively adjusting the number of taps and coefficients based on image characteristics and device capabilities. This allows the system to optimize the balance between processing speed and image quality, achieving acceptable restoration results in mobile devices with limited throughput while minimizing harmful artifacts
Solution Approach 2:
The invention changes the filter parameters (number of taps and coefficients) based on image contrast and frequency content. By dynamically adjusting these parameters, the system achieves high restoration accuracy when possible while maintaining acceptable performance and minimizing ringing and return noise in resource-constrained mobile devices
Data Source
AI summary
An image processing device according to an embodiment of the present invention obtains recovery image data by performing restoration processing using a restoration filter based on a point spread function of an optical system, on original image data obtained from an imaging element by imaging using the optical system. The restoration filter used in this restoration processing (a combination filter and realization filter Fr) is generated by combining multiple base filters Fb. Base filter Fb may be arbitrarily selected from filters that are confirmed beforehand to be effective to prevent image quality degradation.