ROI Picture Encoding With Coefficient Attenuation for XR Bitrate
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding technologies for augmented reality/virtual reality/XR glasses face challenges in achieving a balance between low bitrate and high-quality decoded images, leading to simulator sickness and bulky cabling issues due to high latency and low compression performance.
Innovation Solution
An apparatus and method that utilize the human visual system's spatial resolution properties by attenuating coefficients outside a region of interest, allowing bitrate savings and allocating resources to enhance the region of interest, thereby improving perceived image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If high compression performance is applied to reduce bitrate, then transmission efficiency improves, but image quality deteriorates
Solution Approach 1:
The patent applies different compression strengths to different spatial regions of the image. High-frequency coefficients in the region of interest (fovea) are preserved with minimal attenuation, while coefficients outside the region of interest are strongly attenuated. This local differentiation allows high compression ratios overall while maintaining high image quality in the critical foveal region, resolving the contradiction between bitrate reduction and image quality preservation.
2Manufacturing precision
If high resolution is transmitted to ensure image quality, then decoded picture quality improves, but transmission bandwidth and latency increase
Solution Approach 1:
The patent extracts and transmits only the essential high-frequency information needed for the region of interest (fovea), while discarding or heavily compressing information outside this region. By removing redundant low-quality information from peripheral areas, the system achieves high transmission speed and bandwidth efficiency while maintaining high spatial resolution in the critical foveal region where human vision is most acute.
3Device complexity
If uniform compression is applied to the entire image, then processing complexity is reduced, but perceived image quality deteriorates
Solution Approach 1:
The patent implements spatially adaptive compression by calculating attenuation factors based on the distance of each coefficient from the region of interest. Coefficients within the foveal region undergo minimal attenuation, while those outside the region experience progressive attenuation based on their distance from the center. This local quality differentiation enhances perceived image quality in the critical foveal area without requiring uniform high-complexity processing across the entire image.
Data Source
AI summary
An apparatus for encoding a sample array which signals a spatially sampled representation of a picture, wherein the apparatus is configured for: spectrally decomposing the sample array to obtain a decomposed representation, the decomposed representation comprising a plurality of coefficients, each of which is attributed to one of a plurality of spectral components of the decomposed representation and to one of a plurality of spatial components of the sample array; obtaining, from the plurality of coefficients, a subset of coefficients which are, according to the spatial components to which the coefficients are attributed, located outside of a predetermined region spatially defined within the sample array; attenuating the coefficients of the subset of coefficients, wherein an extent of an attenuation of a coefficient of the subset of coefficients depends on the spatial component to which the coefficient is attributed.