Picture Boundary Extrapolation for Intra Prediction Templates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently handling reference picture boundaries during intra prediction, particularly in extrapolating sample values outside the picture boundary using filter coefficients and intra prediction modes.
Innovation Solution
A method and apparatus for video encoding/decoding that involves determining filter coefficients based on predicted and reconstructed samples within a current picture, and extrapolating values outside the picture boundary using these coefficients, with specific dimensions and templates defined by coding blocks and parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If filter coefficients are determined based on predicted and reconstructed samples within the current picture for extrapolating boundary values, then the accuracy of sample value extrapolation outside the picture boundary is improved, but the computational complexity and processing time increase
Solution Approach 1:
The current picture is divided into multiple templates (first template, second template, third template) with specific dimensional relationships. The second template has dimensions equal to the coding block along the boundary side, while the first template has greater dimensions. This segmentation allows selective application of filtering operations to specific template regions, reducing overall computational complexity while maintaining extrapolation accuracy at the boundary.
Solution Approach 2:
Filter coefficients are pre-determined based on predicted samples from the first template and reconstructed samples from the second template before the actual extrapolation of the third template. This preliminary determination of filter coefficients avoids repeated calculations during the extrapolation process, reducing computational complexity while ensuring accurate boundary value extrapolation.
2Productivity
If the second template dimension along the boundary is set equal to the coding block dimension, then the processing efficiency is improved, but the flexibility in handling different boundary conditions decreases
Solution Approach 1:
The second template is specifically designed with dimensions equal to the coding block along the boundary side, creating a localized processing region optimized for boundary extrapolation. This local quality adjustment allows efficient processing of boundary conditions while the first template with greater dimensions provides additional context when needed, balancing efficiency and flexibility for different boundary scenarios.
Data Source
AI summary
Methods and apparatuses for video decoding and video encoding and a method of processing visual media data are described. The apparatus for video decoding includes processing circuitry configured to determine one of a filter and an intra prediction mode based on reconstructed samples in a first region that are within a current picture. The first region includes a first template and a second template. The second template includes reconstructed samples on a boundary of the current picture and is located between the first template and the boundary of the current picture. The processing circuitry is configured to extrapolate, using the one of the filter and the intra prediction mode, values of samples in a third template that is outside the boundary of the current picture based on values of reconstructed samples in the second template.