Variable Pixel Deblocking Filter for Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding and decoding methods for H.265/HEVC struggle with efficient deblocking filtering, leading to increased data storage requirements and processing complexity.
Innovation Solution
The proposed solution involves an encoder and decoder that determine a filter for deblocking filtering from a set of filters, using a first filter that employs M pixels on both sides of a block boundary and a second filter that uses variable numbers of pixels on each side, allowing for reduced pixel usage and data storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a conventional deblocking filter uses a fixed number of pixels on both sides of the block boundary, then the filtering process is simple and consistent, but the amount of data that must be stored in memory is larger and processing efficiency is reduced
Solution Approach 1:
The patent applies dynamics by making the filter kernel size variable rather than fixed. The filter dynamically adjusts the number of pixels used on each side of the block boundary based on the specific filtering context, allowing the system to optimize between memory usage and processing efficiency for different boundary conditions
Solution Approach 2:
The patent implements local quality by applying different filter configurations to different locations along the block boundary. The filter selects from multiple candidate filters with different pixel counts based on local characteristics, ensuring that each region is processed with the most appropriate filter size for that specific location
2Manufacturing precision
If a deblocking filter uses more pixels for filtering, then the filtering quality is improved, but the data storage requirements and processing complexity increase
Solution Approach 1:
The patent applies partial action by using different numbers of pixels for filtering based on the specific boundary conditions rather than always using the maximum number of pixels. This allows the system to achieve sufficient filtering quality while avoiding the unnecessary processing complexity that would result from always using the largest filter kernel
Solution Approach 2:
The system dynamically selects from multiple candidate filters with different pixel counts based on the local boundary characteristics. This dynamic adaptation allows the filter to achieve high quality where needed while reducing complexity in regions where fewer pixels are sufficient
Data Source
AI summary
A decoder comprises circuitry and memory. The circuitry, using the memory, in operation, determines a number of first pixels and a number of second pixels used in a deblocking filter process, wherein the first pixels are located at an upper side of a block boundary and the second pixels are located at a lower side of the block boundary, and performs the deblocking filter process on the block boundary. The number of the first pixels and the number of the second pixels are selected from among candidates, wherein the candidates include at least 4 and M larger than 4. Response to a location of the block boundary being a predetermined location, the number of the first pixels used in the deblocking filter process is limited to be 4.