Mineralogy Image Blending via Luminosity Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods fail to effectively combine textural and spatial mineral distribution images of a geological sample, leading to a lack of comprehensive understanding of mineralogy, as they either wash out colors or lose valuable information during image blending.
Innovation Solution
A computer-implemented method for blending mineralogy images using a custom blending technique that combines elements from conventional multiply, screen, and overlay modes, weighting pixel values based on luminosity to create a balanced and informative composite image that preserves both high and low intensity areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional image blending methods are used to combine textural and spatial mineral distribution images, then the images can be combined, but valuable information is lost and colors are washed out
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting blend weights based on local image characteristics such as color intensity, luminance, and saturation. Instead of using fixed blending parameters, the system modifies them adaptively to preserve both textural and spatial information while maintaining color fidelity in different regions of the composite image
Solution Approach 2:
The patent implements local quality by applying different blending strategies to different regions of the image based on their specific characteristics. High-intensity regions receive different weighting than low-intensity regions, ensuring that each area contributes optimally to the final composite without washing out colors or losing information
2Measurement precision
If thresholding is applied during image blending to separate mineral regions, then segmentation is improved, but thresholding issues arise that degrade image quality
Solution Approach 1:
The patent avoids thresholding by using continuous parameter-based blending weights derived from color intensity, luminance, and saturation measurements. This approach maintains reliability by eliminating the binary decision-making inherent in thresholding, allowing for smooth transitions and preserving subtle variations in mineral regions
3Loss of information
If one image layer is prioritized over the other during blending, then the dominant layer's information is preserved, but the subordinate layer's information is lost
Solution Approach 1:
The patent uses dynamic parameter adjustment to balance information from both layers. By calculating blend weights based on local color intensity, luminance, and saturation, the system automatically determines the optimal contribution of each layer on a per-pixel basis, preserving information from both textural and spatial images without requiring complex manual intervention
Solution Approach 2:
The blending algorithm performs self-service by automatically adapting to the content of each image region. The system evaluates local characteristics and adjusts blending parameters autonomously, ensuring both layers contribute appropriately without requiring external control or complex preprocessing
Data Source
AI summary
Optimized blending mode for mineralogy images. A luminosity value is determined for a pixel in a base layer or top layer mineralogy image. An image weighting value is determined from the luminosity value and an optional mixing parameter. A multiply value is determined by multiplying the base and top layer pixel values. An overlay value is determined from twice the multiply value if the value of one of the base layer or top layer pixel values is over a threshold, otherwise it is determined by inverting twice the product of the inverted top layer pixel value with the inverted base layer pixel value. A blended image pixel value is determined by adding the multiply value weighted with the image weighting value and the overlay value weighted with the inverted image weighting value.