Intra Block Copy Prediction Beyond CTU Boundaries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing video encoding technologies, such as VVC, constrain the search range of intra picture block compensation (CPR) mode within the current CTU, leading to inefficient memory usage and limited search capabilities.
Innovation Solution
Extend the search range of CPR mode to include parts of the left CTU while maintaining the total memory requirement for storing reference pixels unchanged, utilizing spatial neighboring and non-neighboring blocks for improved video coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the search range of CPR mode is constrained to within the current CTU, then memory usage is controlled, but search capability and coding efficiency are limited
Solution Approach 1:
The patent divides the search space into two parts: the current CTU and a portion of the left CTU. By segmenting the search range in this way, the patent enables extended search capability while maintaining controlled memory usage through selective reference to specific regions rather than the entire left CTU.
Solution Approach 2:
The patent extends the search range from a single CTU boundary to include spatial regions across CTU boundaries by utilizing spatial neighboring and non-neighboring blocks. This dimensional extension allows the encoder to access a broader search space without proportionally increasing memory requirements.
2Productivity
If the search range is extended to include parts of the left CTU, then search capability and coding efficiency improve, but memory requirements increase
Solution Approach 1:
The patent applies local quality by selectively extending the search range only to specific portions of the left CTU that contain spatial neighboring and non-neighboring blocks relevant to the current block. This localized approach improves coding efficiency for blocks that benefit from extended search while avoiding unnecessary memory allocation for regions that do not contribute to coding performance.
Solution Approach 2:
The patent implements partial action by extending the search range only as much as necessary to include useful spatial reference blocks, rather than extending it to cover the entire left CTU. This partial extension achieves the benefit of improved search capability without the full memory cost of a complete CTU extension.
3Measurement precision
If spatial neighboring and non-neighboring blocks are utilized for prediction, then prediction accuracy improves, but computational complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-identifying and organizing spatial neighboring and non-neighboring blocks before the actual prediction process. This preliminary organization of candidate blocks reduces the computational burden during encoding by having reference blocks ready and structured in advance.
Solution Approach 2:
The patent uses copying by referencing previously reconstructed blocks (spatial neighboring and non-neighboring blocks) as templates for predicting the current block. This copying approach improves prediction accuracy by utilizing actual reconstructed pixel data from spatially related blocks rather than generating predictions from scratch.
Data Source
Figure 1
Figure 2A~2B
Figure 2C
AI summary
A method, computer program, and computer system is provided for coding video data. Video data including one or more blocks is received. A current block coded in intra block copy mode or string matching mode is predicted from among the one or more blocks based on a coded block vector or a string offset vector corresponding to one or more spatial neighboring or non-neighboring blocks from among the one or more blocks. The video data, is decoded based on the predicted current block.