Partially Colorized Image Processing with Weighted Color Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image colorization techniques for partially colorized images, such as color-pop images, often lose unique colorful color information, reducing the diversity and details of original colors.
Innovation Solution
An electronic device is designed to identify partially colorized images and automatically colorize the non-emphasized areas while maintaining the emphasized colors, using a method that involves determining weights based on colorfulness indices and synthesizing the original and color images accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If grayscale images are colorized using image colorization techniques including machine learning-based techniques, then the background area can be colored, but unique colorful color information present in color-emphasized areas is lost, reducing the diversity and details of original colors
Solution Approach 1:
The patent segments the image into multiple regions: color-emphasized areas and background areas. By applying different processing strategies to each segment (preserving original colors in emphasized areas while colorizing background areas), the system resolves the contradiction between achieving colorization and preserving color information.
Solution Approach 2:
The patent applies different quality characteristics to different parts of the image. Color-emphasized areas maintain their original color properties (high color fidelity), while background areas receive colorization processing. This local differentiation allows the system to achieve overall colorization while preserving important color information in specific regions.
2Area of stationary object
If the entire image is colorized, then complete color coverage is achieved, but the emphasized colors in specific areas are lost or diluted
Solution Approach 1:
The patent divides the image into color-emphasized areas and background areas, applying different colorization intensities to each segment. This segmentation allows complete color coverage while maintaining color accuracy in emphasized regions by excluding them from aggressive colorization processing.
Solution Approach 2:
Different colorization qualities are applied locally: background areas receive full colorization treatment for complete coverage, while color-emphasized areas maintain their original color properties. This local quality differentiation resolves the contradiction between coverage area and color accuracy.
3Productivity
If colorization is applied to partially colorized images, then non-emphasized areas gain color, but artifacts may be introduced at boundaries between colorized and non-colorized regions
Solution Approach 1:
The patent performs preliminary identification of color-emphasized areas before applying colorization. By pre-segmenting the image and marking protected regions, the system can efficiently apply colorization only where appropriate while preventing artifact formation at boundaries through advance planning of the colorization mask.
Data Source
AI summary
An electronic device may include at least one display, at least one processor, and a memory configured to store instructions. The instructions, when executed by at least one processor, individually and/or collectively may cause the electronic device to: identify whether an input original image is a partially colorized image with an emphasized specific color in a black-and-white background or a monochrome background on the basis of colorfulness, and create a tinted color image in an entire area of the original image; determine a weight corresponding to the original image and a weight corresponding to the color image; and synthesize the input original image and the color image based on the determined weights.