Style-Aware Digital Image Patching for Visual Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital image editing techniques struggle to harmonize the style of patches with the surrounding image, leading to visual inconsistencies when removing defects or objects, as they primarily rely on content similarity without enforcing consistent visual styles.
Innovation Solution
A digital image creation system generates style data and content data for the affected area, selects source images based on both style and content similarity, and transforms the source image's style to match the surrounding area's style using a weighted average of style descriptors, ensuring visually consistent patch incorporation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional systems use an auxiliary image collection (AIC) to generate patches from other digital images, then a larger number of patch candidates are provided, but false positive search results occur that do not match the digital image due to reliance on content alone
Solution Approach 1:
The patent introduces style parameters (color histograms, texture descriptors, edge orientation distributions) alongside traditional content parameters to transform the search criteria. This multi-parameter approach changes the selection basis from content-only to content-plus-style, reducing false positives while maintaining candidate diversity
Solution Approach 2:
The patent combines multiple types of data (content features, style features, color information, texture information) to create a composite matching criterion. This composite approach integrates diverse information sources to achieve both broad candidate selection and precise matching
2Measurement precision
If conventional systems select patches based on content similarity alone, then patches can be found that match the subject matter, but visual style inconsistencies occur between patches and the surrounding image
Solution Approach 1:
The patent adds style parameters (color histograms, texture descriptors, edge orientations) to the traditional content-based matching parameters. This transforms the selection process from content-only to content-plus-style matching, ensuring both subject matter relevance and visual coherence
Solution Approach 2:
The patent applies different matching criteria to different aspects of patch selection: content features for subject matter matching and style features for visual appearance matching. This localized approach to different quality dimensions ensures both content accuracy and style consistency
3Manufacturing precision
If conventional systems copy patches from elsewhere in the same digital image, then style consistency is maintained, but the number of available patch candidates is limited
Solution Approach 1:
The patent creates a universal matching framework that works with both internal patches (from the same image) and external patches (from AIC). The style-based matching mechanism serves as a universal criterion that ensures consistency regardless of the patch source, enabling access to both internal and external candidate pools
Data Source
AI summary
Techniques and systems are described for style-aware patching of a digital image in a digital medium environment. For example, a digital image creation system generates style data for a portion to be filled of a digital image, indicating a style of an area surrounding the portion. The digital image creation system also generates content data for the portion indicating content of the digital image of the area surrounding the portion. The digital image creation system selects a source digital image based on similarity of both style and content of the source digital image at a location of the patch to the style data and content data. The digital image creation system transforms the style of the source digital image based on the style data and generates the patch from the source digital image in the transformed style for incorporation into the portion to be filled of the digital image.