Asymmetric Unsharp Mask Filter for Image Sharpening
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional unsharp mask processing using rotationally symmetrical filters struggles to effectively sharpen images degraded by complex Point Spread Functions (PSF) with asymmetrical aberrations, leading to undershoot and noise issues, particularly on the high frequency side.
Innovation Solution
An image processing method that utilizes a processor to sharpen images by calculating the difference between an input image and a low pass filtered image, where the filters are generated based on the PSF of the optical system, allowing for precise correction and noise minimization through the use of a low pass filter and unsharp mask, which are tailored to the specific image pickup conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a rotationally symmetrical filter is used for unsharp mask processing, then the processing is simple and easy to implement, but it cannot effectively sharpen images degraded by complex asymmetrical PSF, leading to undershoot and insufficient correction
Solution Approach 1:
The patent applies asymmetry by designing the unsharp mask filter to match the asymmetrical characteristics of the optical system's PSF. Instead of using a rotationally symmetrical filter, the filter coefficients are adjusted to reflect the actual asymmetrical blur distribution, enabling accurate correction of asymmetrical aberrations and sagittal halos without causing undershoot.
Solution Approach 2:
The patent implements local quality by making the filter characteristics position-dependent across the image field. The unsharp mask filter is designed with different coefficients for different regions (center vs. periphery) to match the local PSF characteristics at each position, allowing optimal sharpening adaptation to local asymmetrical aberrations.
2Manufacturing precision
If the correction intensity is strengthened to advance the sharpening effect, then the sharpening effect is improved, but noise on the high frequency side particularly increases
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the unsharp mask filter coefficients based on frequency characteristics. The filter is designed to provide stronger correction at frequencies where the PSF degradation is most significant while reducing gain at high frequencies where noise dominates, thus achieving effective sharpening without excessive noise amplification.
Solution Approach 2:
The patent implements feedback by using the known PSF characteristics as a reference to design the unsharp mask filter. The filter coefficients are determined based on feedback from the optical system's actual performance data, allowing the processing to adapt to the specific degradation patterns and avoid amplifying noise in frequency ranges where it would be harmful.
3Adaptability or versatility
If a one-dimensional asymmetrical filter is used as proposed in JP 4618355, then asymmetry in the image height direction is considered, but it cannot improve asymmetry in other directions and the filter asymmetry does not match the PSF blur
Solution Approach 1:
The patent transitions from one-dimensional asymmetrical filtering to two-dimensional unsharp mask filtering. By operating in two dimensions, the filter can simultaneously correct asymmetries in multiple directions (image height and image width) and match the two-dimensional PSF blur distribution more accurately, rather than being limited to correction in a single azimuth direction.
Data Source
AI summary
An image processing apparatus includes a processor that sharpens an input image on the basis a difference between a filter and a low pass filter, which are generated using of information regarding a point spread function of an optical system corresponding to an image pickup condition of the optical system, or a difference between an image obtained by applying the filter to an input image generated by imaging through the optical system and an image obtained by applying the low pass filter to the input image.