Chroma Decoding with Haar Wavelet UV Restoration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Encoding and decoding methods based on neural networks suffer from poor decoding performance and high complexity, particularly in restoring the resolution of UV chroma signals after down sampling, which is computationally intensive and resource-demanding.
Innovation Solution
A low-complexity and efficient post-processing network model framework is proposed to enhance UV signals based on Y signals, utilizing Haar wavelet transforms and feature enhancement through concatenation and residual block networks to improve decoding performance and reduce computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If neural network-based encoding and decoding methods are used to restore UV chroma signal resolution, then decoding performance is improved, but computational complexity and resource requirements increase significantly
Solution Approach 1:
The patent segments the chroma signal processing into multiple frequency bands using Haar wavelet transforms. Instead of processing the entire UV chroma signal at full resolution through complex neural networks, the method divides the signal into different frequency components (LL, LH, HL, HH bands) and processes them separately. This segmentation allows for selective enhancement of important frequency components while discarding less important ones, thereby improving decoding performance with reduced computational complexity.
Solution Approach 2:
The patent extracts and processes only the most important frequency components of the chroma signal. By using Haar wavelet transforms, the method separates the chroma signal into low-frequency (LL) and high-frequency (LH, HL, HH) components. The low-frequency components containing essential visual information are retained and enhanced, while high-frequency components are discarded or simplifiedly processed. This extraction approach improves decoding performance by focusing computational resources on critical signal components.
2Manufacturing precision
If full-resolution UV chroma signals are restored through neural networks, then image quality is improved, but resource requirements and processing time increase
Solution Approach 1:
The patent applies partial action by processing only the essential frequency components of the chroma signal rather than the entire signal at full resolution. Through Haar wavelet decomposition, the method identifies and enhances the low-frequency LL band which contains the most important visual information, while applying simpler processing or discarding the high-frequency bands. This partial processing approach maintains acceptable image quality while significantly reducing processing time and resource requirements.
Solution Approach 2:
The patent applies different processing qualities to different frequency components of the chroma signal. The low-frequency LL band, which contains essential visual information, receives more sophisticated neural network-based enhancement. In contrast, the high-frequency bands (LH, HL, HH) receive simpler processing or are discarded. This local quality approach ensures that computational resources are allocated efficiently, improving image quality where it matters most while reducing overall processing time.
3Measurement precision
If complex neural network models are used for chroma signal enhancement, then decoding accuracy is improved, but device complexity and implementation difficulty increase
Solution Approach 1:
The patent segments the complex chroma signal restoration task into simpler sub-tasks based on frequency bands. By dividing the UV chroma signal into LL, LH, HL, and HH frequency components using Haar wavelet transforms, the method enables the use of simpler neural network models for each band rather than requiring a single complex model to handle the entire signal. This segmentation reduces device complexity while maintaining decoding accuracy through targeted processing of each frequency component.
Data Source
Figure 1~3
Figure 4~5A
Figure 5B~6A
AI summary
The present disclosure provides a decoding method and apparatus, and a device thereof. The decoding method includes: decoding a bit stream corresponding to a current image block to obtain a reconstructed image block, where the reconstructed image block includes a first initial component and a second initial component, where the resolution of the first initial component is greater than or equal to the resolution of the second initial component; generating an adjusted component corresponding to the second initial component based on the first initial component and the second initial component; performing feature enhancement on the adjusted component to obtain a restored target component corresponding to the second initial component. By the technical solution of the present disclosure, decoding performance can be improved, and decoding complexity can be reduced.