Transform Coefficient Decoding With Conditional Dependent Quantization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image decoding and coding technologies, particularly in the context of Versatile Video Coding (Draft 8) and ISO/IEC JTC 1/SC 29/WG 11, 7-17 January 2020, do not effectively address the complexity and cost issues associated with the combination of transform skip, sign data hiding, and dependent quantization, leading to inadequate hardware and software implementation.
Innovation Solution
An image decoding apparatus and apparatus that includes a header decoder to decode flags indicating dependent quantization and transform skip residual quantization, and a TU decoder to perform dependent quantization only when enabled, thereby excluding the combination of transform skip, sign data hiding, and dependent quantization, thus reducing hardware and software implementation costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If transform skip, sign data hiding, and dependent quantization are implemented in image coding schemes, then coding efficiency is improved, but hardware and software implementation costs increase due to system complexity
Solution Approach 1:
The patent extracts and removes the combination of transform skip, sign data hiding, and dependent quantization from the image decoding system. By taking out these complex features that were causing implementation difficulties, the patent simplifies the hardware and software while maintaining adequate coding performance through alternative simpler methods.
Solution Approach 2:
The patent changes the operational parameters of the quantization process by disabling dependent quantization and transform skip features. This parameter change simplifies the implementation while still achieving effective image decoding through modified coding modes that don't require the complex combination of features.
2Adaptability or versatility
If multiple quantization modes (RRC and TSRC) are supported, then adaptability is improved, but device complexity increases due to multiple decoding paths
Solution Approach 1:
The patent removes TSRC mode support from the system, extracting only the necessary RRC mode for image decoding. This eliminates the complexity of maintaining multiple decoding paths while preserving adaptability through flexible parameter control within the simplified RRC framework.
Solution Approach 2:
The patent makes the RRC mode universal by designing it to handle various quantization scenarios without requiring separate TSRC decoding paths. The simplified RRC implementation can adapt to different coding situations through parameter adjustments rather than requiring multiple specialized decoding modes.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image decoding apparatus including a header decoder configured to decode, from coded data, a flag indicating whether dependent quantization is enabled and a flag prohibiting transform skip residual quantization, and a TU decoder configured to decode a transform coefficient in a TU block in an RRC mode in which a LAST position is coded, the LAST position corresponding to a decoding start position for the transform coefficient, or a TSRC mode in which the LAST position is not coded. The TU decoder performs dependent quantization in the RRC mode in a case that the transform coefficient for transform skip is decoded in the TSRC mode, and does not perform dependent quantization in the RRC mode in a case that the transform coefficient for transform skip is decoded in the RRC mode.