CCSO Downsampling Filters for Accurate Chroma Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in effectively reducing reconstruction errors in video data compression, particularly in handling cross-component offset filtering for color components, which affects image quality.
Innovation Solution
Implementing a cross-component offset filtering method that adjusts reconstruction values by using co-located reconstructed samples and neighboring samples of a first color component to derive an offset value for a second color component, enhancing the reconstruction process through adapted sample processing and quantization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If cross-component offset filtering is applied using co-located and neighboring samples, then reconstruction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies downsampling filters to luma samples before the cross-component offset filtering process. This preliminary action reduces the number of luma samples that need to be processed, thereby decreasing computational complexity while still enabling accurate offset calculation for chroma samples. The downsampled luma samples are used to derive offsets that are then applied to reconstruct chroma components with improved accuracy.
Solution Approach 2:
The patent segments the filtering process into distinct stages: first applying downsampling filters to reduce luma sample resolution, then using the downsampled samples to calculate offsets, and finally applying those offsets to chroma samples. This segmentation allows each stage to be optimized independently, balancing computational load with reconstruction quality.
2Productivity
If downsampling filters are applied to luma samples, then processing speed is improved, but sample resolution is reduced
Solution Approach 1:
The patent changes the resolution parameter of luma samples by applying downsampling filters, reducing them from full resolution to a lower resolution suitable for offset calculation. This parameter change accelerates processing speed. The patent then compensates for the resolution reduction by using multiple downsampled samples and their neighboring samples to derive offsets, ensuring that the final chroma reconstruction maintains high precision despite the intermediate downsampling step.
Data Source
AI summary
Various implementations described herein include methods and systems for coding video. In one aspect, a video bitstream includes a current image frame and a first syntax element. An electronic device determines that the first syntax element has a first predefined value indicating that the cross-component sample offset (CCSO) mode is enabled, and generates a set of adapted luma samples including an adapted first luma sample and its adapted neighboring luma samples based on a set of reconstructed luma samples. The reconstructed luma samples include a first luma sample that is collocated with a first color sample. The electronic device determines the first sample offset of the first color sample based on the adapted first luma sample and the one or more adapted neighboring luma samples. The current image frame is reconstructed at least by adjusting the first color sample based on the first sample offset.