Inter Prediction Motion Vector Staging for Lower Encoding Delay

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing encoding and decoding methods in video processing, such as H.265, suffer from significant processing delays.

Innovation Solution

The encoder and decoder utilize memory to derive a first motion vector of a current block using a processed block's motion vector, storing it and then deriving a second motion vector to generate a prediction image, allowing for reduced processing delay by not waiting for the completion of the second motion vector derivation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the decoder waits for the second motion vector to be completed before deriving the first motion vector, then the accuracy of motion estimation is improved, but the processing delay increases

Engineering Contradiction:

Improvemotion estimation accuracyVSAvoidprocessing delay

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by having the decoder derive and store the first motion vector using a processed block's motion vector before the second motion vector is completed. This allows the first motion vector to be prepared in advance, reducing processing delay while maintaining motion estimation accuracy through the use of available motion information from previously processed blocks.

Inventive Principle:

Principle #10Preliminary action

2Device complexity

If the encoder and decoder process motion vectors sequentially, then the processing logic is simplified, but the processing efficiency decreases

Engineering Contradiction:

Improveprocessing logic complexityVSAvoidprocessing efficiency

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The patent segments the motion vector processing into two distinct parts: the first motion vector derived using a processed block's motion vector and stored in memory, and the second motion vector derived separately. This segmentation allows parallel processing of different motion vector components, improving processing efficiency while keeping each segment's logic relatively simple.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces memory as an intermediary component that stores the first motion vector. This intermediary allows the decoding process to proceed without waiting for the second motion vector to be completed, as the first motion vector is already available in memory, thereby improving processing efficiency without significantly increasing logic complexity.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentEP4727131A2Encoding device and encoding method

Publication Date: 2026.04.15 PANASONIC INTELLECTUAL PROPERTY CORP OF AMERICA

Figure 1
Figure 2
Figure 3

AI summary

An encoder (100) includes circuitry (160) and memory (162). Using the memory (162), the circuitry (160): when encoding a current block in an inter prediction mode in which a decoder performs motion estimation (merge mode in S201), derives a first motion vector of the current block (S203); stores, in the memory (162), the first motion vector derived; derives a second motion vector of the current block (S204); and generates a prediction image of the current block by performing motion compensation using the second motion vector (S208). In deriving the first motion vector, the first motion vector of the current block is derived using a first motion vector of a processed block.