Super-resolution loop restoration codec
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital video encoding and decoding technologies face challenges in efficiently managing frame resolution, particularly at low bitrates, leading to suboptimal visual quality and increased computational resources.
Innovation Solution
The implementation of a super-resolution loop restoration framework that allows for upsampling of images within the codec, using loop restoration parameters to enhance visual quality by scaling images from a lower resolution to a higher resolution, specifically focusing on horizontal scaling to reduce computational overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If images are encoded at high resolution, then visual quality is improved, but data size and computational resources increase significantly
Solution Approach 1:
The image is divided into two resolution components: a low-resolution base image that carries most of the data, and a super-resolution component that adds fine details. This segmentation allows the system to transmit less data while maintaining high visual quality through the combination of these components.
Solution Approach 2:
The patent introduces a super-resolution dimension that operates at a different resolution level than the base image. By adding a high-resolution component on top of the low-resolution base image, the system achieves high visual quality without transmitting the entire high-resolution image, thus reducing data size.
2Measurement precision
If loop restoration filtering is applied at super-resolution, then visual quality is enhanced, but computational complexity increases
Solution Approach 1:
The loop restoration filtering is segmented and applied separately to different resolution components. The base image undergoes filtering at its lower resolution, while the super-resolution component is processed independently. This segmentation reduces the computational burden compared to filtering the entire high-resolution image at once.
Solution Approach 2:
Instead of applying full super-resolution filtering to the entire image, the system applies filtering partially to the base image and then combines it with the super-resolution component. This partial action approach maintains visual quality while reducing computational complexity.
3Productivity
If horizontal scaling is performed only in one dimension, then computational overhead is reduced, but image quality may be compromised
Solution Approach 1:
The patent applies asymmetric scaling where only the horizontal dimension undergoes super-resolution transformation, while the vertical dimension remains at the base resolution. This asymmetric approach reduces computational overhead compared to full two-dimensional super-resolution while maintaining adequate image quality for many applications.
Solution Approach 2:
Instead of performing super-resolution scaling in both horizontal and vertical dimensions, the system applies scaling partially only in the horizontal dimension. This partial action reduces computational complexity while still improving image quality sufficiently for the intended application.
Data Source
AI summary
A super-resolution coding mode is described. An encoded image can be decoded from an encoded bitstream stored on a non-transitory computer-readable storage medium. A flag can indicate whether an image was encoded using the super-resolution mode at a first resolution. Responsive to the flag indicating that the image was encoded using the super-resolution mode, bits indicating an amount of scaling of the image are included. The image is decoded from the encoded bitstream to obtain a reconstructed image at the first resolution, and the reconstructed image is upscaled to a second resolution using the amount of scaling to obtain an upscaled reconstructed image. The second resolution is higher than the first resolution. Loop restoration parameters within the bitstream can used for look restoration filtering of the upscaled reconstructed image to obtain a loop restored image at the second resolution.