Arbitrary Block Partitioning for Diagonal-Edge Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video encoding techniques struggle with efficiently encoding pictures that contain edges in directions other than horizontal or vertical, leading to increased data requirements and reduced coding efficiency.
Innovation Solution
The method involves partitioning target blocks into non-rectangular sub-blocks based on edge boundaries, using geometric partitioning, and encoding boundary line information to facilitate intra and inter prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional rectangular block partitioning is used for encoding, then the encoding process is simple and fast, but coding efficiency deteriorates for blocks with complex edges in varying directions
Solution Approach 1:
The target block is segmented into multiple non-rectangular sub-blocks based on detected edge boundaries. Each sub-block is then encoded separately with its own prediction mode, allowing the encoder to adapt to local edge structures and improve coding efficiency for complex textures and diagonal edges.
Solution Approach 2:
The patent introduces arbitrary-shaped non-rectangular blocks beyond the conventional rectangular dimension. By allowing blocks to be defined by arbitrary boundary lines connecting multiple points, the partitioning can follow edge directions in any orientation, not just horizontal or vertical.
2Productivity
If the target block is partitioned into non-rectangular blocks based on boundary lines, then coding efficiency for complex edges is improved, but the amount of boundary line information to be encoded increases
Solution Approach 1:
The patent applies non-rectangular partitioning selectively only to blocks that contain complex edges or textures, while conventional rectangular partitioning is used for simpler blocks. This partial application avoids the overhead of encoding arbitrary boundary information for all blocks, balancing compression gain against bit cost.
3Measurement precision
If intra prediction is performed on non-rectangular blocks, then prediction accuracy for edges in varying directions is improved, but the complexity of determining prediction modes increases
Solution Approach 1:
Different prediction modes are assigned to different non-rectangular sub-blocks based on their local edge characteristics. Each sub-block's prediction mode is determined independently according to its own boundary orientation and texture properties, allowing localized optimization of prediction accuracy.
Data Source
AI summary
A video decoding method for decoding a target block using intra prediction comprises decoding, from a bitstream, boundary line information specifying at least one boundary line for partitioning the target block. The boundary line information allows partitioning the target block into a plurality of non-rectangular blocks. The method further comprises determining intra prediction modes for the non-rectangular blocks based on the boundary line information; generating a prediction block of the target block by performing intra prediction on each of the non-rectangular blocks using the intra prediction modes; reconstructing a residual block of the target block from the bitstream; and reconstructing the target block by adding the prediction block and the residual block.