Intra Prediction Filtering for Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image encoding devices experience significant prediction errors and reduced efficiency due to mismatched edge directions between prediction modes and actual image edges, especially in block sizes other than 8×8 pixels, leading to suboptimal filtering and increased entropy in prediction error signals.
Innovation Solution
A moving image encoding and decoding device that employs an intra prediction unit to generate intermediate predicted values and filters them only at specific positions within a block, using adaptive filtering based on block size, quantization parameters, distance between reference pixels, and intra prediction modes to reduce prediction errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If filtering is applied to all blocks uniformly, then prediction errors are reduced for 8×8 blocks, but prediction efficiency deteriorates for other block sizes due to mismatched filtering strength
Solution Approach 1:
The patent applies different filtering strengths to different block sizes: strong filtering for 8×8 blocks, weak filtering for 4×4 blocks, and no filtering for 16×16 blocks. This local differentiation ensures that each block size receives appropriate filtering treatment, reducing prediction errors without causing mismatched filtering effects.
Solution Approach 2:
The patent changes the filtering parameter (filtering strength) based on block size. By adjusting the filtering parameter according to the specific block size being processed, the system optimizes prediction accuracy for each block type while maintaining adaptability across varying block sizes.
2Manufacturing precision
If strong filtering is applied to reduce prediction errors, then local prediction accuracy improves, but entropy of prediction error signals increases due to over-smoothing
Solution Approach 1:
The patent applies filtering selectively based on local block characteristics rather than uniformly across all blocks. By determining appropriate filtering strength for each block size and applying it locally, the system reduces prediction errors without excessive over-smoothing that would increase entropy.
Solution Approach 2:
The patent applies partial filtering action rather than excessive filtering to all blocks. By applying filtering only where necessary (8×8 blocks with strong filtering, 4×4 blocks with weak filtering, 16×16 blocks with no filtering), the system achieves local prediction accuracy improvement without the entropy increase caused by excessive uniform filtering.
3Productivity
If filtering is applied to all blocks, then prediction efficiency improves, but processing complexity increases due to additional filtering operations
Solution Approach 1:
The patent applies filtering only where necessary based on block size characteristics. By determining that 8×8 blocks benefit from strong filtering, 4×4 blocks from weak filtering, and 16×16 blocks from no filtering, the system improves prediction efficiency for relevant blocks while avoiding unnecessary filtering operations that would increase processing complexity.
Solution Approach 2:
The patent applies partial filtering action rather than filtering all blocks. By selectively applying filtering only to blocks where it provides benefit (8×8 and 4×4 blocks) and omitting it from 16×16 blocks, the system achieves prediction efficiency improvement while minimizing additional processing complexity.
Data Source
AI summary
When carrying out an intra-frame prediction process to generate an intra prediction image by using an already-encoded image signal in a frame, an intra prediction part 4 selects a filter from one or more filters which are prepared in advance according to the states of various parameters associated with the encoding of a target block to be filtered, and carries out a filtering process on a prediction image by using the filter. As a result, prediction errors which occur locally can be reduced, and the image quality can be improved.