Chroma Filter Coefficient Derivation from Luma Geometric Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing adaptive loop filters in video encoding and decoding technologies separately process luma and chroma samples, leading to increased bit rate overheads due to the need to transmit distinct filter coefficients for both, which can be reduced by determining the chroma filter coefficient based on the luma filter coefficient.
Innovation Solution
An image filtering method where the filter coefficient of the chroma filter is determined based on the filter coefficient of the luma filter, allowing only the luma filter coefficient to be transmitted, and using geometrically similar distributions to calculate the chroma filter coefficients from the luma filter coefficients.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If luma and chroma samples are separately processed with distinct filter coefficients, then filtering quality is maintained for both components, but bit rate overhead increases due to transmitting separate coefficients
Solution Approach 1:
The patent merges the separate filter coefficient transmission for luma and chroma into a unified approach where chroma filter coefficients are derived from luma filter coefficients. This combining reduces the number of independent coefficients that need to be transmitted, directly addressing the bit rate overhead issue while maintaining filtering quality through the geometric distribution relationship.
Solution Approach 2:
The patent makes the luma filter coefficients serve a dual purpose: they are used directly for luma filtering and also serve as the basis for generating chroma filter coefficients through geometric distribution. This multi-functionality eliminates the need for separate chroma coefficient transmission, reducing bit rate overhead while preserving filtering effectiveness.
2Reliability
If separate filter coefficients are transmitted for luma and chroma, then filtering effectiveness is maintained, but encoding complexity and transmission data volume increase
Solution Approach 1:
The patent extracts the chroma filter coefficient generation process from the independent transmission stream and derives it from the luma filter coefficients using geometric distribution relationships. This extraction eliminates redundant data transmission while maintaining filtering effectiveness, reducing encoding complexity.
Solution Approach 2:
The patent creates chroma filter coefficients by copying and transforming luma filter coefficients according to geometric distribution patterns. Instead of independently encoding chroma coefficients, the system copies the luma coefficients and applies geometric relationships to generate the chroma version, significantly reducing encoding complexity.
3Loss of information
If unified filter coefficients are used for both luma and chroma, then bit rate overhead is reduced, but filtering precision for chroma may be compromised
Solution Approach 1:
The patent applies local quality by using the same luma filter coefficients for both luma and chroma, but then applies geometric distribution transformations specifically tailored for chroma characteristics. This allows the system to use a unified base (luma coefficients) while maintaining chroma-specific filtering precision through the geometric relationship.
Solution Approach 2:
The patent changes the parameters of the luma filter coefficients by applying geometric distribution relationships to generate chroma-appropriate coefficients. This parameter transformation allows the system to start with luma coefficients but modify them to suit chroma filtering requirements, maintaining precision while reducing overhead.
Data Source
Figure 1~3
Figure 4~5
Figure 6
AI summary
This application provides an image filtering method and apparatus. The method includes: determining a filter coefficient of a chroma filter based on a filter coefficient of a luma filter, where the luma filter and the chroma filter are respectively applied to perform filtering processing on luma samples and chroma samples of a to-be-processed image, all filter coefficients of the luma filter present a first geometric distribution, all filter coefficients of the chroma filter present a second geometric distribution, the first geometric distribution and the second geometric distribution are similar shapes, and a filter coefficient of the chroma filter at an edge of the second geometric distribution is calculated by using a plurality of filter coefficients of the luma filter at an edge of the first geometric distribution; and respectively performing filtering processing on the luma samples and the chroma samples by using the luma filter and the chroma filter.