Selective Prediction Signal Filtering for Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video encoding and decoding techniques suffer from picture quality degradation and visual artifacts due to block-based prediction and quantized block transforms, leading to discontinuities along block boundaries, which reduce the quality of decoded video and the effectiveness of reference frames.
Innovation Solution
The implementation of selective prediction signal filtering, which involves determining performance measurements for different prediction modes and filter combinations, applying filters like Finite Impulse Response (FIR) filters, and selecting the optimal prediction mode and filter to generate a residual for encoding, thereby improving coding efficiency and reducing artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If block-based prediction and quantized block transforms are used, then encoding efficiency is improved, but picture quality degrades and visual artifacts increase
Solution Approach 1:
The patent applies different filtering operations to different regions of the prediction signal based on local characteristics. Edge regions receive different filtering than smooth regions, allowing the system to maintain encoding efficiency while improving picture quality locally where needed, thus resolving the contradiction between global encoding efficiency and local picture quality
Solution Approach 2:
The patent dynamically adjusts filtering parameters and prediction modes based on the content characteristics of each block. By making the filtering process adaptive rather than static, the system can optimize both encoding efficiency and picture quality for different regions without compromising either aspect
2Manufacturing precision
If filtering is applied to prediction samples, then visual artifacts are reduced, but computational complexity increases
Solution Approach 1:
The patent divides the prediction signal into multiple blocks and applies filtering selectively to each block based on its characteristics. This segmentation allows the system to reduce computational complexity by only filtering where necessary, while still improving visual quality in regions that benefit from filtering
Solution Approach 2:
The patent changes filtering parameters such as filter type, strength, and application scope based on the specific characteristics of each prediction block. By adjusting parameters dynamically, the system can achieve good visual quality with minimal computational overhead, resolving the contradiction between quality improvement and complexity reduction
3Measurement precision
If multiple prediction modes and filter combinations are evaluated, then optimal prediction accuracy is achieved, but processing time increases
Solution Approach 1:
The patent evaluates only the most promising prediction modes and filter combinations rather than all possible combinations. By using partial evaluation strategies, the system achieves sufficient prediction accuracy without the time cost of exhaustively searching all options, thus resolving the contradiction between accuracy and processing time
Data Source
AI summary
Disclosed herein are methods and apparatuses for selective prediction signal filtering. One aspect of the disclosed implementations is a method for encoding a frame of a video stream including determining a first performance measurement for a first set of prediction samples identified for a group of pixels using a first prediction mode, generating a filtered set of prediction samples by applying a filter to a second set of prediction samples, wherein at least one of the filtered set of prediction samples or the second set of prediction samples are identified using a second prediction mode, determining a second performance measurement for the filtered set of prediction samples, generating, using a processor, a residual based on the filtered set of prediction samples and the group of pixels if the second performance measurement exceeds the first performance measurement, and encoding the frame using the residual.