Automatic Line Drawing Coloring via Local Style Transfer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing technologies lack the ability to automatically and effectively apply local styles from reference images to line drawings, limiting the ability to create colored images that reflect desired artistic features such as texture, gradation, and pattern.
Innovation Solution
A line drawing automatic coloring program and device that utilize a learned model to extract and apply local styles from reference images to line drawings, using a combination of encoder and decoder networks to generate colored image data, allowing users to designate specific areas for style application and update parameters to minimize loss and achieve accurate coloring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If deep learning is applied to image processing, then image recognition and image generation achieve remarkable results, but the ability to automatically apply local styles from reference images to line drawings remains lacking
Solution Approach 1:
The system segments the reference image into multiple local regions and extracts style features from each region independently. This allows different local styles to be applied to different parts of the line drawing, enabling fine-grained control over the coloring process while maintaining automatic operation.
Solution Approach 2:
The patent introduces an intermediary style extraction mechanism that bridges the reference image and the line drawing. This intermediary process extracts local style features (such as texture, color, and pattern characteristics) from the reference image and applies them to the line drawing, enabling automatic style transfer without requiring manual intervention.
2Manufacturing precision
If local styles are extracted from reference images, then artistic features such as texture, gradation, and pattern can be applied, but the complexity of the processing system increases
Solution Approach 1:
The system extracts only the essential style features (texture, color, gradation, pattern) from the reference image while discarding other irrelevant information. This extraction process simplifies the data that needs to be processed and applied, reducing the overall system complexity while maintaining high coloring accuracy.
Solution Approach 2:
The patent transforms the complex task of style application into a parameter-based process. By representing styles as extractable features (color histograms, texture descriptors, gradient information), the system can manipulate and apply these styles through parameter adjustments rather than complex image processing operations.
3Ease of operation
If users can designate specific places for style extraction and application, then precise control over coloring is achieved, but the operation process becomes more complex
Solution Approach 1:
The system performs preliminary actions by automatically detecting and proposing candidate regions for style extraction based on the reference image content. This preliminary processing reduces the burden on users, as they only need to select from pre-identified candidate regions rather than manually defining all parameters, thus improving ease of operation without significantly increasing interface complexity.
Data Source
AI summary
A line drawing automatic coloring method according to the present disclosure includes: acquiring line drawing data of a target to be colored; receiving at least one local style designation for applying a selected local style to at least one place of the acquired line drawing data; and performing coloring processing reflecting the local style designation on the line drawing data based on a learned model for coloring in which it is learned in advance using the line drawing data and the local style designation as inputs.