Image Processing Apparatus Background Removal Color Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing methods fail to accurately remove backgrounds from color images while preserving color information, especially when the background has large differences in RGB values, leading to degradation in image quality and loss of information in bright or fluorescent colors during monochrome conversion.
Innovation Solution
The execution order of monochrome image generation processing and background removal processing is controlled based on the background removal level, allowing for the selective removal of backgrounds and preservation of color information by applying background removal processing either before or after monochrome conversion, depending on the set level.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If background removal processing is applied to monochromized image data, then background removal is achieved, but color information (bright colors and fluorescent colors) is lost
Solution Approach 1:
The patent applies background removal processing to the original color image data before monochromization processing is performed. By removing the background in advance while the color information is still intact, the system avoids the problem of color information loss that occurs when background removal is applied after monochromization. The extracted background removal level is then used in the subsequent monochromization process to maintain both background removal accuracy and color information preservation.
2Loss of information
If background removal processing is applied to RGB image data before monochromization, then color information is preserved, but backgrounds with large RGB value differences cannot be removed accurately
Solution Approach 1:
The patent performs background removal processing on the original color image data using a background removal level determined from the color image, rather than using fixed thresholds. This approach adapts the background removal criteria to the specific characteristics of the input image, enabling accurate background removal even when backgrounds have large variations in RGB values. The method calculates the background removal level based on the actual distribution of pixel values in the color image, ensuring both accuracy and color information preservation.
3Measurement precision
If conventional background removal is applied to color images, then background removal is achieved, but image quality degrades and color reproducibility is compromised
Solution Approach 1:
The patent performs background removal processing on the original color image data before any monochromization or other processing that could degrade quality. By removing the background in advance while the full color information is preserved, the system achieves accurate background removal without compromising color reproducibility. The background removal level is carefully determined from the color image characteristics to ensure high-quality results.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In a case where a background removal level is less than a threshold, monochrome image data is generated based on color image data and then background removal processing is applied to the generated monochrome image data. In a case where the background removal level is greater than or equal to a threshold, the background removal processing is applied to color image data and then monochrome image data is generated based on the image data having underdone the background removal processing.