Color Gamut Scalability Lookup Table with Variable Chroma Granularity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for color gamut scalability in image coding, such as HEVC, require large buffer sizes for lookup tables to maintain prediction precision, leading to increased device costs, memory resource shortages, and process delays.
Innovation Solution
An image processing device and method that uses a lookup table with granularities of the first and second chroma components rougher than the luma component to predict an image of a second layer with a different color gamut from a first layer, reducing the buffer size while ensuring sufficient prediction precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the lookup table size is increased to maintain prediction precision in color gamut scalability, then the prediction precision is improved, but the buffer size and memory requirements increase significantly
Solution Approach 1:
The patent divides the color gamut conversion process into two stages: first converting from BT.709 to an intermediate color space using a coarse lookup table, then converting from the intermediate space to BT.2020 using another lookup table. This segmentation allows each table to be smaller while maintaining overall prediction precision through the two-step conversion process.
Solution Approach 2:
The patent introduces an intermediate color space (such as XYZ or Lab) as a mediator between the source BT.709 color gamut and the target BT.2020 color gamut. By using this intermediate representation, the system can achieve accurate color gamut conversion with smaller lookup tables compared to direct conversion, reducing memory requirements while preserving prediction quality.
2Measurement precision
If the lookup table buffer size is increased to maintain prediction precision, then the prediction precision is improved, but the device cost increases
Solution Approach 1:
By segmenting the color gamut conversion into multiple stages with smaller intermediate lookup tables, the patent reduces the total memory buffer size required. This directly lowers device cost while maintaining prediction precision through the multi-stage conversion approach.
Solution Approach 2:
The patent changes the parameter representation by introducing an intermediate color space with different numerical characteristics. This parameter transformation allows the use of smaller lookup tables that still achieve the required prediction precision, thereby reducing device complexity and cost.
3Measurement precision
If the lookup table buffer size is increased to maintain prediction precision, then the prediction precision is improved, but process delays occur
Solution Approach 1:
The patent segments the color gamut conversion into smaller processing steps using multiple smaller lookup tables. This segmentation enables more efficient memory access patterns and reduces the time required to complete the conversion process, eliminating process delays while maintaining prediction precision.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided is an image processing device including: a prediction section configured to predict an image of a second layer having a different color gamut from a first layer from an image of the first layer using a lockup table that maps a combination of a luma component, a first chroma component, and a second chroma component of the first layer to a predicted pixel value of the second layer corresponding to the combination. The prediction section uses the lookup table in which granularities of the first chroma component and the second chroma component are rougher than a granularity of the luma component.