Pre-prediction Filtering for Video Block Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards, such as H.265/HEVC, face challenges in reducing noise influence on predicted samples, which can lead to visual artifacts and increased bit-rates, especially in lossy compression scenarios.
Innovation Solution
The implementation of pre-prediction filtering techniques that categorize reference samples into subsets and apply distinct filters, such as smoothing and deblocking filters, to reduce noise impact during the prediction process, allowing for improved reconstruction of image blocks without suppressing picture details.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional prediction techniques are used in H.265/HEVC, then coding efficiency is maintained at acceptable levels, but noise influence on predicted samples increases leading to visual artifacts and higher bit-rates
Solution Approach 1:
The reference samples are divided into multiple subsets based on their spatial location and characteristics. Different filtering operations are applied to different subsets, allowing selective noise reduction while preserving important image details. This segmentation approach enables targeted noise mitigation without uniform filtering of all reference samples.
Solution Approach 2:
Different filtering strengths and types are applied to different regions of reference samples based on local characteristics. Edge regions receive different treatment than smooth regions, preserving local image quality while reducing noise in appropriate areas. This local quality approach ensures that filtering adapts to regional variations in the image content.
2Object-affected harmful factors
If stronger filtering is applied to reference samples, then noise influence is reduced, but picture details may be suppressed
Solution Approach 1:
The filtering operation adapts its strength and type based on local image characteristics such as edge detection and variance analysis. In regions with high detail content, filtering is reduced or skipped to preserve sharpness. In smooth regions with predominant noise, stronger filtering is applied. This local quality approach resolves the contradiction by making filtering intensity spatially variable.
Solution Approach 2:
The filtering parameters are dynamically adjusted based on local image statistics and content characteristics. The filter adapts its behavior in real-time during the prediction process, switching between different filtering modes or intensities depending on the local content. This dynamic adaptation prevents over-filtering of important details while effectively reducing noise where appropriate.
Data Source
AI summary
A prediction technique, which use pre-prediction filtering techniques to reduce noise influence on the predicted samples of a block to be reconstructed, is presented. The prediction techniques suggested herein can be for example used in an encoding apparatus or a decoding apparatus. The reference samples that are used to reconstruct a given block of pixels of an image are categorized or segmented into subsets. The subsets may be processed differently, e.g. the subsets may be subjected to distinct filters. Examples of such filters include smoothing filters and/or deblocking filters, which are applied to the respective subsets of the reference samples.