Video Codec Loop Filter Using Pixel Comparison Lookup Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face challenges in improving video quality during transmission over limited bandwidth channels, particularly in correcting artifacts caused by compression, and require efficient methods to filter pixels effectively across video encoders and decoders.
Innovation Solution
The proposed solution involves constructing a lookup table of offset values based on comparisons between target pixels and neighboring pixels, using equations to derive offset values that adjust pixel values to make them more similar to their neighbors, thereby acting as a low-pass filter, and implementing these operations in both video encoders and decoders to enhance video quality by reducing compression artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional filtering methods are used to correct compression artifacts, then video quality is improved, but computational complexity increases
Solution Approach 1:
The patent transforms the filtering operation from continuous arithmetic computations to discrete parameter lookups. By pre-calculating offset values based on pixel comparison patterns and storing them in lookup tables, the system changes the computational parameters from floating-point arithmetic to integer indexing and addition, significantly reducing computational complexity while maintaining filtering effectiveness.
Solution Approach 2:
The patent performs preliminary computation by pre-calculating offset values for all possible pixel comparison patterns before actual video filtering. These pre-computed values are stored in lookup tables, so during video processing, the system only needs to perform simple comparisons and table lookups rather than complex real-time calculations, thereby reducing operational computational complexity.
2Manufacturing precision
If complex filtering operations are applied to reduce artifacts, then video quality improves, but processing time increases
Solution Approach 1:
The patent changes the computational parameters from multi-step arithmetic operations to single-step table lookups. Each filtering operation is reduced to: (1) compare target pixel with neighbors, (2) form index from comparison results, (3) lookup offset value, (4) add offset to target pixel. This parameter transformation dramatically reduces processing time per pixel.
Solution Approach 2:
By pre-computing offset values for all possible neighbor comparison patterns and storing them in lookup tables, the patent eliminates the need for complex real-time calculations during video processing. The system only needs to perform simple comparisons and retrieve pre-prepared offset values, significantly reducing processing time while maintaining filtering quality.
Data Source
AI summary
Presented herein are techniques for filtering pixels during video coding and decoding operations. Similar operations are performed at a video encoder and a video decoder. For a target pixel in a block of a video frame represented by the encoded bit-stream, a value of the target pixel is compared with neighboring pixels to produce a plurality of comparison results. A particular offset value for the target pixel is derived based on the plurality of comparison results. The target pixel is filtered using the particular offset value. This process is performed for some or all of the pixels of blocks of a video frame.