Loop Filter Boundary Alignment for Video Coding Hardware
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding systems face challenges in reducing line buffer requirements and computational complexity due to data dependencies between Deblocking Filter (DF), Sample Adaptive Offset (SAO), and Adaptive Loop Filter (ALF) processes, leading to increased hardware costs and power consumption.
Innovation Solution
The proposed method aligns loop filter-related boundaries to reduce line buffer requirements by shifting SAO and ALF parameter boundaries, allowing for efficient processing of reconstructed video data within image units, thereby minimizing the need for line buffers and computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If Deblocking Filter, Sample Adaptive Offset, and Adaptive Loop Filter are applied sequentially to enhance picture quality, then video quality is improved, but line buffer requirements and computational complexity increase
Solution Approach 1:
The patent applies Deblocking Filter before SAO and ALF processes to remove block artifacts early in the reconstruction pipeline. By performing deblocking first, the subsequent SAO and ALF operations work on already-smoothed data, reducing the complexity and buffer requirements of later stages while maintaining overall picture quality.
Solution Approach 2:
The patent divides the loop filter processing into distinct sequential stages: Deblocking Filter stage, SAO stage, and ALF stage. Each stage processes specific parameters and boundaries independently, allowing for optimized buffer management and computational efficiency at each segment rather than handling all filters simultaneously.
2Ease of operation
If SAO and ALF parameter boundaries are aligned with image unit boundaries, then processing is simplified, but line buffer requirements increase due to data dependencies
Solution Approach 1:
The patent shifts SAO parameter boundaries and ALF virtual boundaries upward by specific offsets (yS and yV respectively) relative to image unit boundaries. This preliminary boundary adjustment ensures that when processing reaches image unit boundaries, the required filter parameters are already available in the line buffer, eliminating the need to extend buffers further to accommodate parameter dependencies.
Solution Approach 2:
The patent modifies the boundary parameters of SAO and ALF processes by introducing offset values (yS for SAO, yV for ALF). These parameter changes shift the effective processing boundaries upward, allowing the filters to operate with reduced line buffer requirements while maintaining access to all necessary neighboring pixel data for accurate filtering.
3Manufacturing precision
If multiple loop filters are applied to reconstruct video data, then coding artifacts are reduced, but power consumption increases
Solution Approach 1:
The patent performs Deblocking Filter operation first to remove prominent block artifacts early in the reconstruction process. This preliminary deblocking reduces the severity of artifacts remaining for subsequent SAO and ALF stages, allowing these later filters to operate more efficiently with reduced computational requirements, thereby lowering overall power consumption while maintaining video quality.
Data Source
AI summary
A method and apparatus for loop filter processing of reconstructed video data for a video coding system are disclosed. The system receives reconstructed video data for an image unit. The loop filter processing is applied to reconstructed pixels above a deblocking boundary of the current CTU. In order to reduce line buffer requirement and/or to reduce loop filter switching for image units, the sample adaptive offset (SAO) parameter boundary and spatial-loop-filter restricted boundary for the luma and chroma components are determined by global consideration. In one embodiment, the SAO parameter boundary and the spatial-loop-filter restricted boundary are aligned for the luma and chroma components respectively. In another embodiment, the SAO parameter boundary and the spatial-loop-filter restricted boundary for the luma and chroma components are all aligned.