Deblocking Filter for Bit Depth Scalability Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for bit depth scalability in video encoding and decoding, such as tone mapping and simulcast bit-streams, face challenges in compatibility with standards and efficiency, particularly when dealing with the coexistence of 8-bit and 10-bit video systems, and do not effectively address artifacts caused by block boundaries in scalable video coding.
Innovation Solution
The implementation of a deblocking filter at the enhancement layer for bit depth scalability, which adjusts boundary strength based on inverse tone mapping parameters to remove coding artifacts, enabling backward compatibility and efficient adaptation to different bit depths and network bandwidths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If tone mapping methods are applied to convert 10-bit presentation to 8-bit representation, then 10-bit video can be transmitted and displayed on 8-bit devices, but the solution is inherently incompliant with the MPEG-4 AVC Standard's 8-bit profiles
Solution Approach 1:
The video stream is segmented into multiple layers with different bit depths. The enhancement layer carries 10-bit information while the base layer provides 8-bit compatibility, allowing standard-compliant decoding for 8-bit devices while enabling 10-bit reconstruction for capable devices
Solution Approach 2:
The 8-bit base layer is nested within the 10-bit enhancement layer structure. The base layer can be independently decoded for standard compliance, while the enhancement layer adds additional bit depth information when available, creating a nested compatibility structure
2Reliability
If simulcast bit-stream with both 8-bit and 10-bit coded presentations is transmitted, then all current standards are compliant and decoder can choose preferred bit depth, but the solution requires more overhead
Solution Approach 1:
The 8-bit and 10-bit video streams are merged into a single scalable bit-stream where the enhancement layer contains the difference information. This combining approach maintains standard compliance while reducing total bit-stream size compared to transmitting separate simulcast streams
Solution Approach 2:
The bit depth parameter is changed dynamically based on decoder capability. The same bit-stream can be decoded at 8-bit or 10-bit precision depending on the decoder's supported parameters, eliminating the need for multiple separate streams
3Object-affected harmful factors
If deblocking filter is applied at the enhancement layer for bit depth scalability, then coding artifacts caused by block boundaries are removed, but the complexity of the filtering process increases
Solution Approach 1:
The deblocking filter strength is adjusted locally based on the specific block's characteristics and position. Boundary strength values are determined per-block, allowing selective application of filtering where needed while maintaining efficiency in regions where artifacts are less problematic
Solution Approach 2:
The deblocking filter is applied selectively to only those blocks that exhibit significant blocking artifacts. Not all blocks require full filtering strength, so partial application reduces overall complexity while still removing the most noticeable artifacts
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Apparatus and methods are described for encoding and decoding an enhancement layer. A deblocking filter is applied at the enhancement layer for bit depth scalability. The deblocking filter is adjusted to remove coding artifacts caused by local inverse tone mapping for intra-layer texture prediction for the bit depth scalability. The boundary strength of the deblocking filter is adjusted based on a threshold that, in turn, is based on a difference of inverse tone mapping parameters for a block and an adjacent block.