Image Color Cast Correction via Illuminant Information Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image correction technologies face challenges in accurately removing color casts caused by varying illuminants, particularly in scenes with complex lighting conditions and diverse color profiles.
Innovation Solution
The method involves extracting original illuminant information from an input image using an illuminant information extraction model, generating intermediate illuminant information, and determining target illuminant information by combining portions of the original and intermediate information. This target information is then applied to the input image to generate an output image with corrected color casts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing image correction technologies are used to remove color casts, then the processing speed is maintained, but the accuracy of color cast removal deteriorates in complex lighting conditions
Solution Approach 1:
The patent segments the illuminant information into two distinct components: original illuminant information (extracted from the input image) and intermediate illuminant information (generated by the model). This segmentation allows the system to handle different aspects of color correction separately, improving accuracy in complex lighting conditions by combining both extracted and generated illuminant characteristics.
Solution Approach 2:
The patent introduces an intermediate illuminant information as a mediator between the original illuminant extraction and the final white balance adjustment. This intermediate representation allows for more nuanced control over color cast removal, enabling the system to adapt to complex lighting scenarios while maintaining processing efficiency.
2Measurement precision
If a simple white balance algorithm is used, then the processing speed is fast, but the accuracy of color correction deteriorates in diverse color profiles
Solution Approach 1:
The patent performs preliminary extraction of original illuminant information from the input image before applying the main white balance correction. This preliminary action allows the system to prepare accurate color profile data in advance, enabling more precise color correction without requiring overly complex real-time processing algorithms.
Solution Approach 2:
The system uses feedback by combining the extracted original illuminant information with intermediate illuminant information generated by the model. This feedback mechanism allows the algorithm to iteratively refine the color correction process, improving accuracy while maintaining manageable computational complexity through efficient information integration.
3Adaptability or versatility
If only original illuminant information is used for correction, then the processing simplicity is maintained, but the adaptability to different lighting scenarios deteriorates
Solution Approach 1:
The patent merges two sources of illuminant information: the original illuminant extracted directly from the input image and the intermediate illuminant generated by the processing model. This combination enhances adaptability to different lighting scenarios by integrating both direct observation and model-based inference, while the efficient merging process keeps processing complexity manageable.
Data Source
AI summary
An electronic device includes a method performed by an electronic device. The method includes: extracting, from an input image, original illuminant information representing an original illuminant color-cast of the input image, the extracting performed by applying an illuminant information extraction model to the input image; generating, from the original illuminant information, intermediate illuminant information representing an intermediate illuminant color-cast of the input image, the generating performed by applying an illuminant information generation model to the original illuminant information; determining target illuminant information based on the extracted original illuminant information and based on the generated intermediate illuminant information; and generating an output image by applying the determined target illuminant information to the input image.