Adaptive Transform Area Partitioning for Image Coding Efficiency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional image encoding/decoding methods lack the ability to adaptively determine a transform performing area and partition shape of sub transform blocks, limiting their coding efficiency.

Innovation Solution

The method involves determining a transform performing area based on intra-prediction modes and frequency areas, partitioning it into sub blocks using QT or BT partitions, and performing inverse-transforms on these sub blocks, while also adapting the coefficient scan method to improve efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If transform is performed without determining a performing area separately, then the processing is simple, but the coding efficiency is limited

Engineering Contradiction:

Improvecoding efficiencyVSAvoidtransform processing complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The transform block is divided into multiple transform areas based on prediction mode, with different transform coefficients applied to different areas. This segmentation allows the patent to improve coding efficiency by adapting transforms to local characteristics while maintaining manageable processing complexity through systematic division.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different transform coefficients are applied to different transform areas within the same transform block according to local prediction modes. This local quality approach enables the patent to optimize coding efficiency for each region while keeping the overall processing framework relatively simple.

Inventive Principle:

Principle #3Local quality

2Adaptability or versatility

If a fixed transform method is used for all blocks, then the processing is straightforward, but the adaptability to different prediction modes is poor

Engineering Contradiction:

Improveadaptability to prediction modesVSAvoidtransform determination complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The transform method dynamically adapts to different prediction modes by selecting different transform coefficients for different transform areas. This dynamic adaptation improves versatility across various prediction scenarios while maintaining relatively simple processing through predefined coefficient sets for each mode.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent changes transform parameters (coefficients) based on prediction mode to improve adaptability. Different transform coefficients are selected for different prediction modes and transform areas, enhancing versatility while keeping the parameter selection systematic and manageable.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11659174B2Image encoding method/device, image decoding method/device and recording medium having bitstream stored therein

Publication Date: 2023.05.23 IND ACAD COOP GRP OF SEJONG UNIV

AI summary

The present invention provides an image encoding method and an image decoding method. The image decoding method, according to the present invention, comprises the steps of: determining a transformation performing region; segmenting the determined transformation performing region into at least one sub-transform block by using at least one of quadtree segmentation and binary tree segmentation; and performing inverse transformation on the at least one sub-transform block.