MIP Intra Prediction Offset Calculation for Efficient Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Matrix-based Intra Prediction (MIP) mode in video encoding requires significant storage space and increased time for parameter search and invocation due to varying parameters for luma blocks of different sizes, reducing encoding and decoding efficiency.
Innovation Solution
A method where the encoder and decoder determine a MIP block size index based on the current block's size, calculating a first offset using a correspondence relation between pre-stored indexes and offsets, and using this to derive a second offset for predicting block values, thereby simplifying the process and reducing storage and time complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If MIP mode uses different parameters for luma blocks of different sizes, then prediction accuracy is improved, but storage space requirement and parameter search time increase
Solution Approach 1:
The patent applies parameter changes by using a unified offset calculation method that adapts to different block sizes through a single formula involving block width W and height H. Instead of storing separate parameter sets for each block size, the patent dynamically calculates offsets using the relationship: offset = (W - 2) * H / 2, where W and H are the block dimensions. This transforms the problem from storing multiple parameter sets to using a single parameter set with dynamic calculation, reducing storage requirements while maintaining accuracy for various block sizes.
2Reliability
If MIP mode stores parameters for various block sizes, then prediction performance is maintained, but encoding and decoding efficiency decreases
Solution Approach 1:
The patent extracts the essential parameter (offset) from the complex parameter sets that would normally be stored for each block size. By identifying that the offset can be universally calculated using the block dimensions W and H through the formula offset = (W - 2) * H / 2, the patent removes the need to store and search through multiple parameter sets. This extraction simplifies the parameter management process, improving encoding and decoding efficiency while maintaining prediction performance through accurate offset calculation.
Data Source
AI summary
Provided are a picture encoding and decoding method, an encoder, and a decoder. The decoder decodes a bitstream to obtain a size and a coding mode of a current block; when the coding mode of the current block is a Matrix-based Intra Prediction (MIP) mode, determines a value of a MIP block size index corresponding to the current block according to the size of the current block, calculates a second offset based on the MIP block size index, a constant value and reconstructed values of adjacent pixels corresponding to the current block; determines a first prediction value of the current block according to the second offset; and determines a reconstructed value of the current block based on the first prediction value.