Guide-Image Color Mapping for Fast Video Colorization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image and video color adjustment processes are time-consuming and costly, typically requiring intensive manual frame-by-frame coloration to achieve desired color alterations.
Innovation Solution
A computer-implemented method and system that utilizes machine learning and lookup tables (LUTs) to automatically adjust image colorization by generating color mapping logic based on guide images, allowing for efficient conversion between different dynamic ranges and formats, such as from standard dynamic range (SDR) to high dynamic range (HDR), and between different definitions like standard definition (SD) to high definition (HD) and ultra-high definition (UHD, e.g., 4K).
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual frame-by-frame coloration process is used, then color quality and accuracy are improved, but time consumption and cost increase significantly
Solution Approach 1:
The patent replaces the manual mechanical color adjustment process with an automated computer-based system that uses machine learning models and lookup tables to perform color mapping and conversion operations, eliminating the need for frame-by-frame manual intervention while maintaining color quality
Solution Approach 2:
The patent creates lookup tables that store pre-computed color mapping relationships derived from training data, allowing the system to quickly copy and apply color transformation patterns across multiple frames without repeating the complex analysis process for each frame individually
2Productivity
If automated color adjustment systems are implemented, then productivity is improved, but color accuracy and quality may deteriorate
Solution Approach 1:
The patent performs preliminary color mapping analysis and lookup table generation during an offline training phase using representative training images, so that the actual processing of video frames can proceed quickly by simply applying the pre-computed mappings without performing complex analysis in real-time
Solution Approach 2:
The patent introduces lookup tables as an intermediary structure that bridges the gap between the complex color analysis model and the fast processing requirement, storing pre-computed color transformation data that can be quickly referenced and applied during video processing without requiring real-time computation
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer-implemented method of adjusting colorization of one or more input images of a plurality of input images is provided. The method comprises receiving the plurality of input images; receiving a guide image corresponding to at least one selected input image of the plurality of input images, wherein the guide image comprises additional colorization relative to the at least one selected input image; generating a color mapping logic configured to adjust a color of the at least one selected input image to match the additional colorization of the guide image; outputting an indication of the color mapping logic; and generating one or more output images by applying, based upon the indication of the color mapping logic, the color mapping logic to one or more additional input images of the plurality of input images.