Video Coding Prediction Offset for Residual Signal Quantization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques, such as those in the HEVC standard, introduce coding artifacts during the quantization process, leading to impairments in reconstructed video data, which can be alleviated but not fully resolved by existing in-loop processing methods like deblocking filters and sample adaptive offset.
Innovation Solution
A prediction offset is derived from the reconstructed residual signal of coded blocks to calculate a new residual signal at the encoder, reducing the absolute values of the final residual signal and thereby reducing the bits required for encoding, and this offset is applied during both encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If quantization is applied during video encoding, then compression ratio is improved, but coding artifacts are introduced degrading picture quality
Solution Approach 1:
The patent applies prediction offset derivation as a preliminary action before quantization. By deriving the prediction offset from reconstructed residual signals and applying it to the current residual signal before quantization, the method prepares the data in advance to minimize quantization artifacts. This preliminary intervention allows the quantization process to operate on pre-corrected data, reducing the harmful effects of quantization while maintaining compression efficiency.
Solution Approach 2:
The patent implements a feedback mechanism where the prediction offset is derived from previously reconstructed residual signals. The reconstructed residual signal from prior blocks is fed back into the system to compute the prediction offset, which is then applied to the current residual signal. This feedback loop allows the system to learn from previous reconstruction errors and continuously improve prediction accuracy, thereby reducing coding artifacts while maintaining compression.
2Object-affected harmful factors
If deblocking filter and sample adaptive offset are applied, then picture quality is improved, but processing complexity increases
Solution Approach 1:
The patent applies prediction offset derivation as a preliminary action before the reconstructed signal is stored in the reference picture buffer. By deriving and applying the prediction offset at this early stage, the method improves picture quality proactively before the signal becomes reference data for future predictions. This preliminary correction reduces the need for aggressive post-processing filtering, thereby lowering overall processing complexity while maintaining or improving picture quality.
Solution Approach 2:
The prediction offset acts as an intermediary between the residual signal and the reconstructed signal. Instead of directly applying complex post-processing filters like deblocking and SAO to correct artifacts, the method introduces the prediction offset as an intermediate correction layer. This intermediary approach simplifies the overall processing chain by addressing artifacts earlier in the pipeline, reducing the burden on subsequent filtering stages.
3Loss of substance
If motion compensation prediction is used, then temporal redundancy is exploited improving compression, but reference picture reconstruction accuracy deteriorates due to accumulated artifacts
Solution Approach 1:
The patent implements feedback by deriving the prediction offset from reconstructed residual signals of previously coded blocks. This feedback mechanism allows the system to continuously correct for accumulated quantization artifacts in reference pictures. By feeding back the reconstructed residual information and using it to compute prediction offsets for subsequent blocks, the method compensates for degradation in reference picture accuracy, maintaining higher precision over extended encoding sequences.
Solution Approach 2:
The patent applies prediction offset derivation as a preliminary correction before reference pictures are used for future motion compensation predictions. By deriving and applying the offset before the reconstructed signal is stored in the reference picture buffer, the method proactively prevents artifact accumulation from propagating to future reference pictures. This preliminary intervention maintains reference picture accuracy without requiring re-encoding or complex post-processing.
Data Source
AI summary
A method and apparatus for video encoding and decoding with prediction offset in a video coding system are disclosed. A prediction offset for a current block is derived from a reconstructed residual signal of one or more coded blocks. At the encoder side, a residual signal for the current block is calculated by subtracting a prediction signal and the prediction offset from an original signal of the current block, and the residual signal is encoded to generate an encoded bitstream. At the decoder side, a residual signal for the current block is obtained from input data, a reconstructed signal for the current block is calculated by adding a prediction signal and the prediction offset to the residual signal, and the current block is decoded using the reconstructed signal.