Shading Correction for Virtual Slide Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Virtual slide technology faces issues with brightness unevenness in images due to illumination and optical system non-uniformity, leading to unnatural boundaries when stitching images, which existing shading correction methods fail to adequately address.
Innovation Solution
An image processing device that acquires first and second image groups with overlapping subject areas, calculates normalized and non-normalized shading components based on luminance ratios, and performs shading correction using these components to address brightness unevenness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images are acquired for shading correction, then shading correction precision is improved, but the number of images required increases and processing time increases
Solution Approach 1:
The patent segments the shading correction process into two distinct components: normalized shading components (obtained by dividing image luminance by flat area luminance) and non-normalized shading components (obtained by dividing adjacent image luminance). This segmentation allows the system to use different components for different correction scenarios, improving precision while reducing the number of images needed compared to traditional methods that require multiple calibration images.
Solution Approach 2:
The patent performs preliminary calculation of shading components from a small set of images (only two images: one with flat area and one adjacent image). These pre-calculated normalized and non-normalized shading components are then reused for correcting multiple images, eliminating the need to acquire and process multiple calibration images for each correction operation.
2Reliability
If traditional shading correction methods are used, then brightness unevenness is addressed, but natural boundaries are created in stitched images and periodic patterns appear
Solution Approach 1:
The patent applies different shading correction approaches to different regions of the image. Normalized shading components (based on flat areas) are applied to regions where uniform reference is appropriate, while non-normalized shading components (based on adjacent images) are applied to regions requiring contextual continuity. This local differentiation eliminates artificial boundaries and periodic patterns that occur when a single correction method is applied uniformly across the entire image.
Solution Approach 2:
The patent introduces shading components as an intermediary element that mediates between the raw image data and the final corrected image. By calculating normalized and non-normalized shading components from overlapping image regions and using these as correction factors, the system smoothly transitions luminance values across image boundaries, eliminating unnatural boundaries and periodic patterns in stitched images.
3Measurement precision
If shading correction is performed using existing methods, then some brightness correction is achieved, but high-precision correction in short time is not realized
Solution Approach 1:
The patent performs preliminary calculation of normalized and non-normalized shading components from just two images (one containing a flat area and one adjacent image). These pre-calculated components are then reused for correcting multiple images, achieving high-precision shading correction without the time-consuming process of acquiring and processing multiple calibration images for each correction operation.
Solution Approach 2:
The patent changes the approach from acquiring multiple calibration images to calculating shading components through mathematical operations on luminance ratios. By transforming the correction process into parameter calculation (normalized shading component = image luminance / flat area luminance; non-normalized shading component = adjacent image luminance / flat area luminance), the system achieves high precision with reduced processing time.
Data Source
AI summary
An image processing device includes: an image acquiring unit that acquires first and second image groups, each image sharing a common part of a subject with the other image in different first and second directions; a calculating unit that calculates, as a shading component, a ratio of luminance of an area in one or more images, the area sharing a common subject with another area including a flat area whose shading component is constant in a single image other than the one or more images, to luminance of the another area, for each of the first and second image groups; and a correcting unit that performs a shading correction on shading areas in the images using the shading component. The shading component includes a normalized shading component based on luminance of the flat area, and a non-normalized shading component based on luminance of an area other than the flat area.