Focus-Stacked Image Processing for Smooth Images and Accurate Height Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing all-in-focus image generation methods and height maps suffer from discrepancies between calculated luminance values and actual heights due to influence from surrounding pixel positions, leading to inaccuracies in analysis.
Innovation Solution
An image processing method that calculates luminance values and composite heights by considering image reference values and sharpnesses at both the current and surrounding pixel positions, using Gaussian weighting for distance and sharpness, to generate a smooth all-in-focus image and appropriate height map.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If the luminance value at each pixel position is calculated by reflecting surrounding pixel positions, then the all-in-focus image becomes smoother, but the height map becomes inconsistent with the actual all-in-focus image
Solution Approach 1:
The patent applies different processing strategies to different regions: for pixels where all surrounding pixels have the same image reference value, the original image reference value is used (maintaining local accuracy); for pixels with mixed image reference values, a weighted average is calculated (achieving local smoothing). This local differentiation resolves the contradiction between smoothness and accuracy.
Solution Approach 2:
The patent introduces a new parameter 'composite height' that combines multiple image reference values with different weights. By changing from a single image reference value to a composite parameter, the system achieves both smoothing (through averaging) and accuracy preservation (through selective weighting), resolving the contradiction.
2Reliability
If multiple picked-up images are combined to generate all-in-focus image, then the image quality improves, but the complexity of image processing increases
Solution Approach 1:
The patent performs preliminary classification of pixels into two groups (those with uniform surrounding image reference values and those with mixed values) before the actual composite calculation. This preliminary action simplifies the overall processing by dividing the complex task into two straightforward sub-tasks, reducing processing complexity while maintaining image quality.
Solution Approach 2:
The patent segments the image processing into distinct steps: calculating sharpness for each picked-up image, determining image reference values, classifying pixels into groups, and calculating composite heights. This segmentation breaks down the complex processing into manageable stages, reducing overall complexity while ensuring high image quality.
Data Source
AI summary
In an image processing method, a plurality of sharpnesses are calculated at each pixel position from a plurality of picked-up images which are acquired while a focal position is changed along an optical axis, and by comparing the plurality of sharpnesses at each pixel position with one another, an image reference value to be referred for determination of a luminance value at the pixel position of an all-in-focus image is determined. The luminance value at each pixel position of the all-in-focus image is calculated on the basis of image reference values in a pixel position group including the pixel position and surrounding pixel positions of the pixel position. A composite height at each pixel position is acquired on the basis of image reference values and/or sharpnesses in a pixel position group including the pixel position and the surrounding pixel positions, and a height map is thereby generated.