Arbitrary Block Partitioning for Edge-Adaptive Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression techniques struggle to efficiently encode pictures with edges in varying directions, leading to increased data requirements and reduced coding efficiency due to the use of square or rectangular block-based encoding.
Innovation Solution
The method involves partitioning target blocks into non-rectangular sub-blocks based on edge boundaries, allowing for efficient encoding and decoding of video data by using arbitrary block partitioning techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If square or rectangular block-based encoding is used, then the encoding process is simple and systematic, but coding efficiency decreases and data requirements increase for pictures with edges in varying directions
Solution Approach 1:
The patent divides the picture into multiple blocks and further partitions each block into sub-blocks based on edge detection results. This segmentation allows the encoding process to adapt to local image characteristics, particularly edges in varying directions, thereby improving coding efficiency while maintaining systematic processing through hierarchical division.
Solution Approach 2:
The patent introduces dynamic block partitioning where the partitioning structure changes based on detected edge directions. Instead of using fixed square or rectangular blocks throughout, the system dynamically adjusts block shapes and partitioning patterns to match local image features, resolving the contradiction between systematic encoding and coding efficiency.
2Ease of operation
If square or rectangular block-based encoding is used, then the encoding structure is regular and easy to implement, but the amount of encoding data increases
Solution Approach 1:
The patent applies different encoding strategies to different regions of the picture based on local edge characteristics. Blocks containing edges in varying directions are partitioned into sub-blocks with directional encoding, while regions without significant edges use standard encoding. This local adaptation reduces the overall amount of encoding data needed while maintaining regularity in the overall encoding structure.
3Ease of manufacture
If conventional block partitioning is used, then processing is straightforward, but image enhancement effect is insufficient for complex textures
Solution Approach 1:
The patent implements dynamic partitioning where block structures adapt to local image content. For regions with complex textures and edges, the system dynamically creates directional sub-block partitions that follow edge orientations, enhancing image representation precision while maintaining straightforward processing through automated detection and adaptation.
Solution Approach 2:
The patent changes encoding parameters such as block size, partitioning depth, and transformation types based on detected image characteristics. By adjusting these parameters dynamically according to edge density and direction, the system achieves better image enhancement effects for complex textures while keeping the base processing framework straightforward.
Data Source
AI summary
A video decoding apparatus for decoding a target block using intra prediction includes a processor configured to: obtain, from a bitstream, boundary line information specifying at least one boundary line for partitioning the target block and prediction mode information, wherein the boundary line information allows partitioning the target block into a plurality of non-rectangular blocks; determine intra prediction modes of the non-rectangular blocks using the prediction mode information; generate a prediction block of the target block by performing intra prediction on each of the non-rectangular blocks using the determined intra prediction modes; reconstruct a residual block of the target block from the bitstream; and reconstruct the target block by adding the prediction block and the residual block. Determining intra prediction modes of the non-rectangular blocks includes: constructing a candidate list of intra prediction mode candidates; and selecting an intra prediction mode of the non-rectangular block from the candidate list.