Color-Space Transform Bit-Depth Equalization for Screen Content

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional video coding techniques are inefficient for compressing computer-generated screen content with discrete tones and sharp lines, as they assume continuous tones and smoothness, leading to sub-optimal coding performance and increased decoder complexity due to inadequate color-space transforms and transform matrix derivation methods.

Innovation Solution

The implementation of a color-space conversion process for video blocks using techniques like in-loop color-space transform and adaptive color-space transforms, such as YCoCg and YCoCg-R, to equalize bit-depths of luma and chroma components, reducing decoder complexity and improving coding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional video coding techniques are used for screen content, then coding is based on assumptions of continuous tones and smoothness, but this leads to sub-optimal coding performance and increased decoder complexity due to inadequate color-space transforms

Engineering Contradiction:
Improvecoding efficiencyVSAvoiddecoder complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent applies parameter changes by modifying the bit-depth of chroma components to match the luma component bit-depth through bit-shifting operations. This allows the color-space transform to operate on equalized bit-depths, improving coding efficiency for screen content while maintaining manageable decoder complexity through systematic parameter adjustment.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent segments the color components into luma and chroma parts, applying different bit-depth equalization strategies to each. The chroma components are specifically processed with bit-shifting to match luma precision, allowing targeted optimization without affecting the entire video data structure uniformly.

Inventive Principle:
Principle #1Segmentation

2Productivity

If color-space transform is applied with different bit-depths for luma and chroma components, then the transform can be performed, but this results in signal decline and requires Delta QPs compensation

Engineering Contradiction:
Improvecoding performanceVSAvoidsignal decline
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent creates equipotentiality by equalizing the bit-depth of chroma components to match the luma component before applying the color-space transform. This is achieved through right-shifting chroma values by a calculated number of bits (bdChroma - bdLuma), ensuring all components operate at the same precision level, thereby preventing signal decline and eliminating the need for Delta QPs compensation.

Inventive Principle:
Principle #12Equipotentiality

3Reliability

If conventional transform methods are used, then standard video coding processes are maintained, but this leads to inadequate color-space transforms for screen content with discrete tones and sharp lines

Engineering Contradiction:
Improvecoding performance for screen contentVSAvoidtransform matrix derivation
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent modifies the transform parameters by equalizing bit-depths through bit-shifting operations on chroma components. This parameter change enables the color-space transform to work effectively on screen content with discrete tones and sharp lines, improving reliability without complicating the transform matrix derivation since the same YCoCg-R matrix can be used with adjusted input precision.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3117610B1Modifying bit depths in color-space transform coding
Publication Date: 2024.02.28 QUALCOMM INC
  • EP3117610B1 patent drawingFigure 1
  • EP3117610B1 patent drawingFigure 2
  • EP3117610B1 patent drawingFigure 3

AI summary

In general, this disclosure describes techniques for coding video blocks using a color-space conversion process. A video coder, such as a video encoder or a video decoder, may determine a bit-depth of a luma component of the video data and a bit-depth of a chroma component of the video data. In response to the bit-depth of the luma component being different than the bit depth of the chroma component, the video coder may modify one or both of the bit depth of the luma component and the bit depth of the chroma component such that the bit depths are equal. The video coder may further apply the color-space transform process in encoding the video data.