Image Color Segmentation for Multi-Color Fabric Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Standard spectrophotometers fail to effectively measure colors in multi-color fabrics, prints, and materials with overlapping colors or indistinct boundaries, leading to difficulties in color separation and analysis.
Innovation Solution
A system and method using a processor to segment colors in images by selecting color values, comparing pixel values, averaging those within a pre-determined range, and outputting averaged color data, allowing for precise color separation and reduced computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard spectrophotometers are used to measure fabric color, then solid color fabrics can be measured effectively, but multi-color fabrics and prints cannot be measured due to the averaging effect
Solution Approach 1:
The patent segments the image into multiple color regions by comparing each pixel's color values to reference color values and grouping pixels with similar colors. This segmentation allows distinct color regions to be identified and measured separately, resolving the averaging effect problem of traditional spectrophotometers on multi-color fabrics.
2Adaptability or versatility
If image capture devices are used to analyze multi-color material, then color separation becomes possible, but separating colors in images with overlapping areas or indistinct boundaries remains difficult
Solution Approach 1:
The patent changes the parameter threshold for color classification by using a predetermined range (threshold value) to determine if pixel color values match reference colors. This parameter adjustment allows flexible handling of indistinct boundaries and overlapping colors, enabling effective color separation in complex images.
3Measurement precision
If all pixel color values are processed individually for color separation, then accurate color identification is achieved, but computational complexity increases significantly
Solution Approach 1:
The patent merges pixels with similar color values into color groups or regions. By comparing pixels to reference color values and grouping those within a predetermined range, the system processes groups of pixels rather than individual pixels, significantly reducing computational complexity while maintaining color identification accuracy.
4Device complexity
If color values are averaged for all pixels in an image, then computational complexity is reduced, but color separation accuracy is lost due to the averaging effect
Solution Approach 1:
The patent performs segmentation before averaging by first grouping pixels into distinct color regions based on their color values. Then, averaging is applied separately to each segmented region rather than to all pixels together. This approach maintains color separation accuracy while still reducing computational complexity through localized averaging.
Data Source
AI summary
Systems, methods and computer program products for segmenting colors in a sample image comprised of pixels using a processor configured by code executing for selecting at least one color value present within the sample image and for each color selected from the sample image comparing the color values of each pixel of the sample image to the selected color value and averaging the color values of each pixels of the sample image having a color value difference below a pre-determined threshold. The method also includes outputting the average color value as a color value data object; and associating each averaged color value with sample image.