Image Reconstruction with Adaptive Prediction Granularity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The existing image coding methods face inefficiencies in prediction due to the use of a unified prediction method that does not adapt to varying content within coding units, leading to low prediction efficiency.

Innovation Solution

The method involves acquiring a quantization parameter and prediction mode for a current coding block, adjusting the prediction granularity accordingly, and performing prediction based on the adjusted granularity to obtain a prediction sample block, followed by reconstructing the sample block using the quantization parameter.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If a unified prediction method is used for the entire coding unit, then the hardware implementation is simple and the processing is efficient, but the prediction accuracy deteriorates when different contents exist in the coding unit

Engineering Contradiction:
Improveprediction method simplicityVSAvoidprediction accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The coding unit is divided into multiple prediction units based on content characteristics. Each prediction unit can be processed with appropriate prediction methods, allowing the system to handle different contents within the coding unit separately, thereby improving prediction accuracy while maintaining reasonable complexity

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different prediction methods or granularities are applied to different regions within the coding unit based on local content characteristics. This allows the prediction process to adapt to local variations in the image data, improving overall prediction accuracy without requiring complex global processing

Inventive Principle:
Principle #3Local quality

2Productivity

If the prediction granularity is fixed, then the processing speed is high and the hardware complexity is low, but the prediction efficiency deteriorates when adapting to different content regions

Engineering Contradiction:
Improveprocessing speedVSAvoidcontent adaptation capability
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The prediction granularity is made dynamic rather than fixed. The system can adjust the prediction granularity based on the content characteristics of different regions, allowing faster processing for simple regions and more detailed processing for complex regions, thus balancing speed and adaptability

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The prediction granularity parameter is changed adaptively based on content analysis. By modifying this parameter according to the actual content requirements, the system achieves both high processing speed for simple cases and high prediction efficiency for complex cases

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4679829A1Image reconstruction method and apparatus, device, and storage medium
Publication Date: 2026.01.14 HANGZHOU HIKVISION DIGITAL TECHNOLOGY CO LTD
  • EP4679829A1 patent drawingFigure 1~2
  • EP4679829A1 patent drawingFigure 3
  • EP4679829A1 patent drawingFigure 4

AI summary

The present application belongs to the technical field of image coding and discloses an image reconstruction method and apparatus, device, and storage medium. The image reconstruction method includes: acquiring a quantization parameter and a prediction mode corresponding to a current coding block; adjusting a prediction granularity corresponding to the prediction mode; predicting according to the adjusted prediction granularity to obtain a prediction sample block; obtaining a reconstructed sample block based on the prediction sample block and the quantization parameter.