Image Processing Device Combining Images for High Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image-processing technologies face challenges in suppressing artifacts caused by subject movement and position displacement when combining images captured with a sensor shifting method, especially in situations with low pixel value differences or subtle movements, leading to noise masking and reduced image resolution.
Innovation Solution
An image-processing device and method that calculates position displacement distributions and correlation levels between comparative images with different color filters, adjusting combining ratios to correct and enhance image resolution by combining standard and reference images in a high-resolution space, using interpolation and filtering techniques to minimize artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images are combined by shifting the image sensor to achieve high resolution, then image resolution is improved, but artifacts such as ghosts occur in regions where the imaging subject is moved
Solution Approach 1:
The patent changes the parameter used for determining combining ratios from simple similarity calculation to correlation level calculation based on position displacement distribution. This parameter change enables more accurate assessment of image correspondence despite subject movement, resolving the contradiction between achieving high resolution through multiple image combination and suppressing artifacts caused by subject movement.
2Object-generated harmful factors
If combining ratios are controlled based on similarities among images, then artifacts due to movement are suppressed, but noise masking occurs in regions with low pixel value differences
Solution Approach 1:
The patent substitutes the mechanical similarity calculation approach with a correlation level calculation approach that incorporates position displacement distribution. This substitution replaces the flawed mechanism that caused noise masking with a more robust mechanism that can distinguish between actual subject movement and noise, thereby suppressing artifacts without causing information loss.
3Object-generated harmful factors
If the correlation level is calculated based on position displacement distribution, then artifact suppression is improved, but calculation complexity increases
Solution Approach 1:
The patent performs preliminary calculation of position displacement distribution between comparative images before calculating correlation levels. This preliminary action organizes the displacement data in advance, making the subsequent correlation level calculation more efficient and manageable, thus reducing overall calculation complexity while maintaining improved artifact suppression.
Data Source
AI summary
An image-processing device includes a high-resolution combining portion that generates a combined image by combining a standard image and at least one reference image, which are acquired by capturing images of a subject in time series by using an imaging element in which multiple types of color filters are arrayed in individual pixels, in a high-resolution space where the resolution is greater than those of the standard image and the reference images. The image-processing device also includes a position-displacement-distribution calculating portion that calculates a distribution of position displacements between two comparative images that are individually formed of pixels corresponding to the different types of color filters in the generated combined image, a correlation-level calculating portion that calculates, for individual pixels, correlation levels between the two comparative images based on the calculated distribution, and an image correcting portion that corrects the combined image based on the calculated correlation levels.