Video Block Prediction Using Power-of-Two Reference Pixels

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The inefficiency of video encoding due to the long operation cycle of division operations in obtaining an average value of reference pixel values for predicting encoding blocks, leading to slow calculation speed and low encoding efficiency.

Innovation Solution

Perform prediction using target reference pixels whose quantity is the e-th power of a target system, replacing division operations with shift operations to calculate the pixel average value, thereby avoiding division and enhancing calculation speed and encoding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If division operation is used to calculate the average value of reference pixel values, then the predicted value can be obtained, but the operation cycle is long resulting in slow calculation speed

Engineering Contradiction:
Improveprediction accuracyVSAvoidcalculation speed
Core Design Contradiction:
Measurement precisionVSSpeed

Solution Approach 1:

The patent changes the parameter of reference pixel quantity from an arbitrary number to a power of 2 (e.g., 2, 4, 8, 16 pixels). This parameter change enables the use of shift operations instead of division operations, as dividing by a power of 2 is equivalent to a right shift operation in binary arithmetic, thereby resolving the contradiction between maintaining prediction accuracy and improving calculation speed

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent substitutes the mechanical division operation with a shift operation in the computational system. Since shift operations are fundamentally simpler and can be executed in a single clock cycle compared to multi-cycle division operations, this substitution directly addresses the speed problem while preserving the mathematical equivalence of the average value calculation

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If division operation is used to obtain pixel average value, then the predicted value is calculated, but the operation cycle is long leading to low encoding efficiency

Engineering Contradiction:
Improvepixel average value accuracyVSAvoidencoding efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent modifies the parameter of reference pixel count to be a power of 2, which transforms the division operation into a shift operation. This change maintains the accuracy of pixel average value calculation while dramatically improving encoding efficiency, as shift operations are much faster and can be parallelized more effectively in hardware implementations

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces the computationally intensive division operation with a simple shift operation in the video encoding system. This substitution reduces the operational complexity and cycle time, directly enhancing encoding productivity without sacrificing the precision of pixel average value computation

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3869805B1Method and apparatus for video encoding, method and apparatus for video decoding, computer device and storage medium
Publication Date: 2025.12.03 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • EP3869805B1 patent drawingFigure 1
  • EP3869805B1 patent drawingFigure 2~3
  • EP3869805B1 patent drawingFigure 4A~4B

AI summary

This application relates to a video encoding method and apparatus, a video decoding method and apparatus, a computer device, and a storage medium. The video encoding method includes: obtaining a current encoding block to be encoded in a current video frame; determining target reference pixels corresponding to the current encoding block, a target quantity corresponding to the target reference pixels being the eth power of a target system, e being a positive integer, the target system being a system used for calculating a predicted value of the current encoding block; obtaining a predicted value corresponding to the current encoding block according to the target reference pixels; and performing video encoding on the current encoding block according to the predicted value, to obtain encoded data. According to the foregoing method, encoding efficiency can be improved.