Adaptive Sampling in Video Deblocking Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding methods require significant computation resources and memory due to fixed sampling in deblocking filtering, which is inefficient for high spatial resolution and high frame rate video content.
Innovation Solution
A locally adaptive sampling method is applied in the in-loop filtering stage of video signal coding, determining an adaptive sampling rate or pattern based on picture properties like block size and parameters, allowing for different sampling rates or patterns between blocks and using rate conversion values to optimize sampling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed sampling is used in deblocking filtering, then the filtering process is simple, but computation resources and memory requirements increase significantly
Solution Approach 1:
The patent applies different sampling rates to different regions (blocks) of the video signal based on their specific characteristics. Instead of using a uniform fixed sampling rate across the entire image, the system determines an adaptive sampling rate for each block, allowing blocks with smoother regions to use lower sampling rates and blocks with detailed regions to use higher sampling rates. This local adaptation reduces overall computation resources while maintaining filtering quality.
Solution Approach 2:
The patent introduces dynamic sampling rate adjustment where the sampling rate is not fixed but adapts based on block properties such as texture complexity, edge density, and frequency characteristics. The system dynamically determines the optimal sampling rate for each block during the deblocking filtering process, enabling the filtering algorithm to adjust its behavior according to the local signal characteristics and reduce computational burden in regions that do not require high sampling rates.
2Device complexity
If fixed sampling is used in deblocking filtering, then the processing is straightforward, but memory requirements and access bandwidth increase
Solution Approach 1:
The patent implements local adaptive sampling where each block is processed with a sampling rate tailored to its specific characteristics. Blocks with simple textures or smooth regions use lower sampling rates, reducing the number of pixels that need to be stored and accessed in memory. Conversely, blocks with complex textures or fine details use higher sampling rates. This localized approach reduces overall memory requirements and memory access bandwidth while maintaining processing simplicity through automated rate determination.
3Productivity
If adaptive sampling is applied to all pixels uniformly, then coding efficiency improves, but computation complexity increases
Solution Approach 1:
The patent applies adaptive sampling selectively and locally to different blocks based on their visual characteristics rather than uniformly to the entire image. The system analyzes block properties such as texture complexity, edge density, and frequency content to determine the optimal sampling rate for each block. This localized adaptation achieves compression efficiency improvements in regions that benefit from it while avoiding the computational complexity burden in regions where simple fixed sampling suffices.
Solution Approach 2:
The patent applies adaptive sampling only where necessary (partial action) rather than universally. By determining the sampling rate based on block characteristics, the system applies high sampling rates only to blocks that require them for quality preservation, while using lower sampling rates for blocks where they are not needed. This partial application of adaptive sampling achieves the benefits of improved compression efficiency without the full computational complexity cost of uniform adaptation across the entire image.
Data Source
AI summary
Disclosed herein is a method of performing a deblocking filtering on a video signal. The method may include determining an adaptive sampling rate or an adaptive sampling pattern based on the property information of a decoded picture and performing the deblocking filtering using samples to which the adaptive sampling rate or the adaptive sampling pattern has been applied. The property information of the decoded picture may include at least one of a block size and a picture parameter.