Video Block Transform Derivation From L-Shaped Surrounding Pixels

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video encoding and decoding technologies face inefficiencies in transform index signaling, particularly in advanced standards like VVC, where the need for more transform types increases signaling overhead without providing additional benefits.

Innovation Solution

The method and apparatus determine transform indices at the decoder side by analyzing L-shaped surrounding pixels of a block, eliminating the need for explicit signaling and allowing for a higher number of transforms, thereby improving coding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If more transform types are added to increase coding efficiency, then the adaptability of transform selection is improved, but the signaling overhead increases

Engineering Contradiction:

Improvetransform selection adaptabilityVSAvoidsignaling overhead

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The decoder derives the transform index autonomously by analyzing L-shaped surrounding pixels of the current block without requiring explicit signaling from the encoder. This self-service mechanism eliminates the need for additional transform index signaling bits while maintaining adaptability in transform selection based on local block characteristics

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The transform index derivation is performed preliminarily at the decoder side using pre-defined rules based on L-shaped pixel patterns before actual transform application. This preliminary determination of transform type avoids the need for subsequent signaling and enables efficient transform selection based on local image characteristics

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20250365419A1Methods and apparatuses for encoding/decoding a video

Publication Date: 2025.11.27 INTERDIGITAL CE PATENT HOLDINGS SAS

AI summary

A method comprising decoding a video is provided, wherein decoding the video includes determining at least one transform for a block of the video based on at least one part of reconstructed pixels surrounding the block, decoding the block by applying the at least one determined transform. An apparatus for decoding the video is also provided. Corresponding method and apparatus for encoding a video are also provided.