JPEG AI Encoding with Region-Based Quality Matrix Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing JPEG AI image coding standards struggle with spatial dimension quality adjustment for different regions of interest and background regions, necessitating improved bit rate allocation.
Innovation Solution
Introduce a target quality matrix representing image quality for each region, scaling residual maps and Gaussian distribution parameters to adjust bit rates, enabling spatial dimension quality adjustment through JPEG AI encoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If JPEG AI encoding is applied to improve compression efficiency, then compression efficiency is improved, but spatial dimension quality adjustment capability deteriorates
Solution Approach 1:
The patent introduces a target quality matrix where different quality values are assigned to different spatial regions of the image. This allows region-of-interest (ROI) areas to maintain high quality while background areas use lower quality, achieving spatial dimension quality adjustment within JPEG AI encoding framework.
Solution Approach 2:
The patent modifies the JPEG AI encoding process by incorporating quality matrix parameters and region identification parameters. These parameter changes enable the encoder to adaptively adjust compression strength across different spatial regions, resolving the contradiction between compression efficiency and quality control flexibility.
2Ease of manufacture
If uniform quality encoding is applied to all regions, then encoding simplicity is maintained, but bit rate allocation efficiency deteriorates
Solution Approach 1:
The patent segments the image into different regions (ROI and non-ROI) using region identification parameters. This segmentation allows the encoding process to apply different quality levels to different segments, improving bit rate allocation efficiency while maintaining manageable encoding complexity through systematic region classification.
Data Source
Figure 1a
Figure 1b
Figure 2
AI summary
Embodiments of this application disclose an encoding method and apparatus and a decoding method and apparatus, and relate to the field of media technologies, so that spatial dimension quality adjustment can be performed on image content through JPEG AI. The method includes: obtaining a target quality matrix; scaling a first residual map and/or first Gaussian distribution parameter information based on the target quality matrix to obtain a second residual map and/or second Gaussian distribution parameter information; and generating a bitstream based on the target quality matrix, the second residual map, and/or the second Gaussian distribution parameter information. The target quality matrix represents image quality of each region in a residual map of a feature domain, the first residual map is the residual map of the feature domain, and the first Gaussian distribution parameter information is Gaussian distribution parameter information of the residual map of the feature domain.