Temporal Motion Vector Predictor Derivation for Image Compression

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The increasing demand for high-resolution and high-quality images has led to higher data transmission and storage costs, necessitating more efficient image compression techniques.

Innovation Solution

A method for deriving a temporal motion vector predictor is proposed, which involves selecting a reference picture, determining a prediction block, and deriving a temporal motion vector predictor from the motion information of the reference prediction unit, using specific storage unit blocks and scanning priorities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If high-resolution and high-quality images are transmitted and stored, then image quality is improved, but transmission costs and storage costs increase

Engineering Contradiction:

Improveimage qualityVSAvoiddata amount

Core Design Contradiction:

Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent segments the image into blocks and divides motion information into storage unit blocks (SUBs) within each block. This segmentation allows selective processing and prediction at different granularities, improving compression efficiency while maintaining image quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary motion vector prediction by determining temporal motion vector predictors (TMVPs) before final encoding. Motion information is pre-processed and stored in SUBs, enabling efficient inter-picture prediction and reducing the data needed for transmission and storage.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If motion information is stored for each block, then prediction accuracy is improved, but storage requirements increase

Engineering Contradiction:

Improveprediction accuracyVSAvoidstorage requirement

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent merges motion information from multiple blocks into storage unit blocks (SUBs). Each SUB contains motion information that represents multiple adjacent blocks, reducing the total number of separate storage entries while maintaining prediction accuracy through representative motion vectors.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent changes the granularity parameter of motion information storage from block-level to SUB-level. By adjusting this parameter, the system achieves a balance between prediction accuracy and storage requirements, storing motion information at an optimized resolution that reduces data量 while maintaining effectiveness.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250193434A1Method for encoding image and non-transitory computer readable storage medium storing a bitstream generated by a method

Publication Date: 2025.06.12 LG ELECTRONICS INC

AI summary

The method for deriving a temporal motion vector predictor according to the present invention comprises the steps of: selecting a reference picture for a current block; deciding a predictor block corresponding to a predetermined storage unit block, as a 5 reference prediction unit for the current block, in the reference picture; and deriving the temporal motion vector predictor from motion information of the decided reference prediction unit. The present invention enhances image compression efficiency.