Image Capture Restoration Filter Switching for Focus State Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image capture devices face challenges in restoring moving images to high-resolution quality without overcorrection, especially when images are captured during focus adjustment, leading to lower image quality due to inappropriate restoration processing.
Innovation Solution
An image capture device with an auto focus adjustment mechanism that uses two filters - a first filter for in-focus images and a second filter with weaker restoration strength for defocused images, along with an in-focus determination section to select the appropriate filter based on the image's focus state, ensuring accurate restoration processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If restoration processing is performed on defocused images using a restoration filter generated from focused image PSF/OTF, then restoration processing can be applied to all captured images, but image quality deteriorates due to overcorrection and inappropriate restoration
Solution Approach 1:
The patent changes the restoration filter parameters based on the focus state of the image. When defocus is detected, the system switches from a restoration filter generated from focused image PSF/OTF to a restoration filter generated from defocused image PSF/OTF, thereby adapting the restoration parameters to match the actual image conditions and preventing quality deterioration
Solution Approach 2:
The patent introduces dynamic switching between different restoration filters based on real-time focus detection. The system continuously monitors image focus state and dynamically adjusts the restoration processing by selecting appropriate filters, making the restoration process adaptive rather than static
2Manufacturing precision
If a restoration filter with fine point spread is used to prevent overcorrection, then image disturbance is suppressed, but high-resolution moving image restoration becomes difficult to achieve
Solution Approach 1:
The patent generates different restoration filter parameters based on focus state. For defocused images, it uses PSF/OTF characteristics specific to defocused conditions rather than focused conditions, allowing optimal restoration strength and resolution enhancement tailored to each image's actual state
3Manufacturing precision
If auto focus adjustment is performed during moving image capture, then sharp images can be obtained, but time difference occurs until the subject is brought into focus, resulting in inappropriate restoration processing
Solution Approach 1:
The patent performs preliminary focus detection and determines the focus state before applying restoration processing. By detecting defocus condition in advance and selecting appropriate restoration filters accordingly, the system ensures that restoration is always applied appropriately regardless of the focus adjustment timeline
Solution Approach 2:
The patent implements a feedback mechanism where the focus detection result is fed back to the restoration processing section. This closed-loop control ensures that the restoration filter selection is continuously adjusted based on the actual focus state, preventing inappropriate restoration during focus transitions
Data Source
AI summary
A first restoration processing section 110 and a second restoration processing section 120 perform restoration processing on images (luminance data Y), which are successively captured by an image capture section, using a first filter 102, which is a restoration filter generated corresponding to a point spread function of an optical system, and a second filter 104 of which a restoration strength is weaker than that of the first filter 102. Depending on a result of determination which is input from an in-focus determination section 150 and indicates whether or not the image at the current time point is in a target in-focus state, a selection section 122 selects and outputs either luminance data YA which is processed using the first filter 102 or luminance data YB which is processed using the second filter 104.