Color Correction Model Training for Multi-Standard Video Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing display technologies face challenges in accurately reproducing colors across different video standards due to variations in display characteristics, leading to inefficiencies in color correction and resource-intensive look-up tables for each standard.
Innovation Solution
A method and apparatus for generating a color correction model through an initial color correction model training process, utilizing a first and second color space conversion, reference point determination, and interpolation algorithms to create a trained model capable of correcting colors across multiple video standards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate color correction models are developed for different video standards, then color accuracy for each standard is improved, but device complexity and development resources increase
Solution Approach 1:
The patent develops a universal color correction model that can handle multiple video standards (BT.709, BT.2020, DCI-P3, etc.) through a single unified framework. The model uses standard color space conversion procedures and interpolation algorithms that work across different video standards, eliminating the need to develop separate models for each standard while maintaining color accuracy.
Solution Approach 2:
The patent changes the approach from creating multiple fixed models to a single adaptive model that adjusts its parameters based on the input video standard. By modifying the color space conversion parameters and interpolation points according to the specific video standard being processed, the model achieves high color accuracy for different standards without increasing structural complexity.
2Adaptability or versatility
If multiple separate color correction models are maintained for different video standards, then resource occupation increases, but color correction capability across standards is improved
Solution Approach 1:
The patent creates a single color correction model that serves multiple video standards, reducing resource occupation from storing multiple separate models to maintaining one universal model. The model processes different video standards by adjusting conversion parameters and interpolation algorithms rather than requiring separate model instances for each standard.
3Ease of manufacture
If conventional color correction methods are used, then implementation is simple, but color consistency across different display characteristics is poor
Solution Approach 1:
The patent introduces an intermediary color space conversion process that transforms colors from different video standards into a common reference frame before applying correction. This intermediary step ensures color consistency across different display characteristics by providing a standardized intermediate representation that can be uniformly processed regardless of the source video standard.
Solution Approach 2:
The patent employs iterative optimization with feedback mechanisms where the color correction model is trained using sample pixels and their corresponding theoretical output data. The model continuously adjusts its parameters based on the difference between actual and expected color outputs, improving color consistency across different video standards and display characteristics through this feedback-driven refinement process.
Data Source
AI summary
A method of generating a color correction model, including: acquiring first color coordinates of a sample pixel in a first color space; converting the first color coordinates of the sample pixel to second color coordinates of the sample pixel in a second color space; inputting the second color coordinates of the sample pixel into an initial color correction model to generate sample output data; and training the initial color correction model according to the sample output data and theoretical output data corresponding to the first color coordinates, to obtain a trained color correction model, where the trained color correction model is configured to perform color correction on a target pixel in each of a plurality of video standards.