Selective Digital Image Correction via Metadata Layers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital image processing systems that automatically correct red-eye issues can incorrectly identify objects, leading to customer dissatisfaction, and lack the ability for users to selectively edit automatic corrections without reanalyzing the original image.
Innovation Solution
A system that allows users to selectively edit automatic corrections by generating a low-resolution preview image with metadata representing the corrected areas, enabling users to view and compare original and corrected images without reprocessing the high-resolution image, and includes an optional learning algorithm to improve future corrections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If automatic red eye correction is applied to the displayed image, then red eye issues are corrected, but the system cannot selectively edit corrections without reanalyzing the original image
Solution Approach 1:
The patent segments the correction process by creating separate correction layers for different correction types (red eye correction layer, cosmetic correction layer, etc.). Each layer can be independently manipulated, allowing users to selectively edit or remove specific corrections without affecting others or requiring reprocessing of the entire original image.
Solution Approach 2:
The system performs preliminary analysis of the original image to identify correction areas and stores this information in metadata. This preliminary action enables subsequent selective editing operations to be performed quickly by simply modifying the stored correction data rather than reprocessing the entire image.
2Adaptability or versatility
If the system generates a preview image with metadata representing corrected areas, then user control and comparison are enabled, but the system complexity increases
Solution Approach 1:
The patent introduces metadata as an intermediary layer between the original image and the corrected display. This metadata stores correction information without modifying the original image data, allowing versatile user control and comparison while keeping the core image processing system relatively simple.
Solution Approach 2:
The system creates a copy of the original image with corrections applied as a separate layer, rather than permanently modifying the original. This copying approach enables users to view and compare original and corrected versions while maintaining system simplicity through non-destructive editing.
3Ease of operation
If multiple correction layers are stored separately, then selective editing is enabled, but the quantity of data stored increases
Solution Approach 1:
The patent applies local quality by storing correction data only for specific regions of the image where corrections are needed, rather than storing correction information for the entire image. This region-specific storage approach enables selective editing while minimizing data storage requirements.
Data Source
AI summary
A method of allowing for selective editing of a digital image that has undergone a correction of at least one area of interest is provided. The method includes: providing for the identification and correction of the at least one area of interest in the digital image while maintaining the digital image; providing the correction of the at least one area of interest as correction metadata; providing a low resolution corrected image using a low resolution version of the digital image and the correction metadata; and allowing the correction metadata to be edited to establish modified correction metadata, wherein a revised version of the low resolution corrected image includes the modified correction metadata.