Non-Square Block Reference Sample Filtering in Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing still image or moving image compression technologies face challenges in determining whether to perform reference sample filtering and selecting the appropriate filter type for non-square blocks, leading to ambiguity in criteria and inefficient processing.
Innovation Solution
A method and apparatus that determine whether to perform reference sample filtering and select the filter type based on parameters related to the current block and its peripheral block, including edge lengths, sample numbers, quantization parameters, and distance between prediction and reference samples, allowing for independent filtering of top and left reference samples without requiring separate tables for non-square blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If reference sample filtering is performed using existing square block tables, then memory consumption is reduced, but prediction accuracy for non-square blocks deteriorates due to ambiguous criteria
Solution Approach 1:
The patent changes the parameters used for filtering determination from block size alone to a combination of intra prediction mode and block shape (square vs. non-square). By introducing block shape as a new parameter dimension, the system can differentiate filtering requirements for non-square blocks while still using the existing square block tables, thus maintaining memory efficiency while improving prediction accuracy.
2Measurement precision
If separate tables are created for non-square blocks, then prediction accuracy improves, but device complexity increases
Solution Approach 1:
The patent makes the existing square block tables multi-functional by introducing a block shape parameter. The same tables can serve both square and non-square blocks, with the filtering determination process adapting based on the block shape. This eliminates the need for separate tables while maintaining improved prediction accuracy for non-square blocks.
3Ease of operation
If filtering criteria are simplified for non-square blocks, then ease of operation improves, but prediction accuracy deteriorates
Solution Approach 1:
The patent applies different filtering determination rules based on the local characteristic of block shape. For non-square blocks, the system specifically checks both the intra prediction mode and the block shape condition, while square blocks follow the traditional size-based criteria. This localized adaptation ensures simplicity in determination while maintaining high prediction accuracy for each block type.
Data Source
AI summary
Disclosed are an image processing method and an apparatus therefor. Specifically, the image processing method comprises the steps of: obtaining an intra prediction mode; obtaining a reference sample using a neighboring sample of the current block; determining whether to perform filtering on the reference sample; filtering the reference sample if it is determined to perform the filtering; and generating a prediction block using the reference sample or the filtered reference sample on the basis of the prediction mode, wherein the current block is a non-square block, the reference sample comprises a left reference sample and a top reference sample, and whether to perform the filtering is determined on the basis of at least one of a current block parameter or a surrounding block parameter.