In-Loop Filter Offset Scaling for Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image compression technologies face challenges in enhancing encoding/decoding efficiency, particularly in reducing distortion, error, and artifacts like blocking and ringing effects, especially with the increasing demand for high-resolution and high-quality image formats such as UHD and HDR.
Innovation Solution
A method and apparatus that adaptively obtain a parameter for scaling an offset from a bitstream or neighboring blocks, using a scale residual value to derive a scale value for the in-loop filter, allowing for efficient pixel value adjustment in the encoding and decoding process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If an in-loop filter is applied to reduce distortion and artifacts in reconstructed images, then image quality is improved, but device complexity and computational overhead increase
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the offset scale value based on neighboring block characteristics. The offset parameter is scaled using a scale value derived from neighboring blocks, allowing the filter strength to adapt to local image features. This resolves the contradiction by optimizing image quality through parameter adaptation while avoiding uniform application of complex filtering across all blocks.
Solution Approach 2:
The patent implements local quality by applying different offset scaling to different blocks based on their neighboring blocks' characteristics. Each block's offset is scaled independently using locally derived scale values, allowing the in-loop filter to adapt to local image features such as edges and textures. This localized approach improves image quality where needed while reducing unnecessary processing in uniform regions.
2Productivity
If adaptive offset scaling is performed using neighboring block information, then compression efficiency is improved, but computational complexity increases
Solution Approach 1:
The patent applies preliminary action by deriving and storing offset parameters and scale values from neighboring blocks before processing the current block. The neighboring blocks' offset parameters are used to predict and pre-compute scale values, which are then applied to the current block. This preliminary computation improves compression efficiency by reducing the need for complex real-time calculations during main processing.
Solution Approach 2:
The patent uses copying by reusing offset parameters and scale values from neighboring blocks. Instead of computing unique parameters for each block, the method copies and adapts parameters from already-processed neighboring blocks, significantly reducing computational complexity while maintaining compression efficiency through adaptive scaling.
3Adaptability or versatility
If offset parameters are derived from bitstream or neighboring blocks, then encoding flexibility is improved, but decoding complexity increases
Solution Approach 1:
The patent applies universality by designing a dual-source parameter derivation mechanism that can obtain offset parameters from either the bitstream or neighboring blocks based on availability and efficiency considerations. This multi-functional approach provides encoding flexibility to choose the optimal source while maintaining a unified decoding process that handles both cases through the same offset scaling mechanism.
Data Source
AI summary
The present invention provides a method and apparatus for encoding/decoding an image and for adjusting a reconstructed pixel by scaling an offset of an in-loop filter having a scaled offset value. The method includes: obtaining information of a scale default value of a current block from a bitstream; obtaining offset parameter merge information of the current block from the bitstream; obtaining the offset parameter of the current block on the basis of the offset parameter merge information; determining whether a scale residual value of the current block is obtained from the bitstream on the basis of the offset parameter merge information; obtaining a scale value on the basis of the scale default value and the scale residual value; scaling an offset on the basis of the offset parameter and the scale value; and adjusting a pixel value of the current block on the basis of the scaled offset.