Video Encoding In-Loop Filter Parameter Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video codecs face inefficiencies in encoding and decoding high-resolution video content due to limitations in encoding methods, particularly in calculating and applying filter parameters, leading to increased circuit size and power consumption.
Innovation Solution
The proposed solution involves predicting Sample Adaptive Offset (SAO) and Adaptive Loop Filter (ALF) filter parameters before deblocking filtering, allowing for entropy encoding and reducing waiting times for data to be processed, thereby improving encoding speed and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If filter parameters are calculated and applied after deblocking filtering, then filtering accuracy is improved, but encoding speed deteriorates and circuit complexity increases
Solution Approach 1:
The patent performs SAO and ALF filter parameter calculations before deblocking filtering, rather than after. This preliminary action allows the parameters to be ready in advance, enabling parallel processing and reducing the critical path delay, thus improving encoding speed while maintaining filtering accuracy through subsequent refinement steps.
Solution Approach 2:
The patent divides the filtering process into separate stages: parameter calculation stage, deblocking filtering stage, and in-loop filtering stage. This segmentation allows each stage to be optimized independently and processed in parallel where possible, reducing overall processing time and circuit complexity while maintaining accuracy.
2Manufacturing precision
If filter parameters are calculated and applied after deblocking filtering, then filtering accuracy is improved, but circuit size and power consumption increase
Solution Approach 1:
By calculating filter parameters before deblocking filtering, the patent enables earlier initialization of filter structures and reduces the need for large buffers and complex timing control logic that would be required if calculations were performed after deblocking. This reduces overall circuit size and power consumption.
Solution Approach 2:
The patent uses prediction mechanisms where filter parameters for current blocks are copied or predicted from previously processed blocks, reducing the need for complete recalculation and thereby reducing circuit complexity and power consumption while maintaining adequate filtering accuracy.
3Manufacturing precision
If filter parameters are calculated and applied after deblocking filtering, then filtering accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs filter parameter calculations in advance before deblocking filtering, allowing these computations to overlap with other processing operations. This eliminates sequential waiting time and reduces total processing time while maintaining filtering accuracy through subsequent refinement.
Solution Approach 2:
The patent enables continuous processing by allowing parameter calculations to proceed continuously without waiting for deblocking to complete. This continuous action eliminates idle time and keeps processing pipelines full, thereby reducing overall processing time while maintaining accuracy.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
Provided is a video encoding apparatus comprising an encoder configured to generate encoded video data by encoding an input video; a decoder configured to decode video data to which a filter to compensate for a pixel value is to be applied, wherein the video data is from the encoded video data; an adaptive loop filter (ALF) parameter predictor configured to generate an ALF filter parameter by using information of the decoded video data, wherein the ALF filter parameter is configured to be applied to an ALF filter to compensate for a value of a current pixel by using a value of a neighboring pixel adjacent to the current pixel and a filter coefficient with respect to the neighboring pixel; a sample adaptive offset (SAO) filter unit configured to apply a SAO filter to the decoded video data, wherein the SAO filter compensates for a value of a current pixel by using at least one of an edge offset and a band offset; an ALF filter unit configured to apply, by using the ALF filter parameter, the ALF filter to video data to which the SAO filter has been applied; and an entropy encoder configured to perform entropy encoding on the ALF filter parameter.