Video Decoding Sample Spacing for Size-Mismatched Reference Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding and decoding methods do not adequately consider the size differences between current and reference images, leading to suboptimal reconstruction quality.
Innovation Solution
The proposed methods and apparatuses compare the sizes of current and reference images to select and process reference samples using n-tap filters based on size comparisons, adjusting filter coefficients and sample selection to improve reconstruction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If reference samples are selected from a reference image with different size than the current image, then the ability to handle size variations is improved, but the reconstruction quality deteriorates due to improper sample spacing
Solution Approach 1:
The patent changes the parameter of sample spacing by introducing an interval value that is determined based on the size relationship between the current image and reference image. When the reference image has a different size, the interval is adjusted accordingly to ensure proper sampling density, thereby maintaining reconstruction quality while adapting to size variations.
Solution Approach 2:
The patent makes the sample selection process dynamic by determining the interval value dynamically based on the size comparison between current and reference images. This dynamic adjustment allows the system to adapt to different image size scenarios while maintaining optimal reconstruction quality through proper sample spacing.
2Manufacturing precision
If an n-tap filter is applied to reference samples, then the reconstruction accuracy is improved, but the processing complexity increases
Solution Approach 1:
The patent changes the parameter of filter application by conditionally applying the n-tap filter based on the size relationship between images. The filter is applied only when necessary (when reference image size differs from current image size), thereby improving reconstruction accuracy for size-mismatched scenarios while avoiding unnecessary processing complexity when image sizes are equal.
Solution Approach 2:
The patent makes the filtering process dynamic by controlling whether the n-tap filter is applied based on real-time size comparison results. This dynamic conditional application allows the system to adjust processing complexity according to the specific scenario, applying filters only when needed for accurate reconstruction.
3Loss of time
If the interval between reference samples is increased, then the processing time is reduced, but the reconstruction quality deteriorates
Solution Approach 1:
The patent changes the parameter of sample interval by determining the interval value based on the size relationship between current and reference images. This conditional parameter adjustment ensures that the interval is optimized for each scenario, reducing processing time when appropriate while maintaining reconstruction quality through proper sampling density.
Solution Approach 2:
The patent makes the sample interval dynamic by adjusting it according to the size comparison results. This dynamic adjustment allows the system to optimize the balance between processing time and reconstruction quality by selecting appropriate intervals for different image size scenarios.
Data Source
AI summary
An image decoding method includes: comparing a size of a current image including a current block with a size of a reference image; selecting, in the reference image, reference samples corresponding to current samples in the current block according to a motion vector of the current block; and reconstructing the current block based on the reference samples, wherein, when the size of the reference image is greater than the size of the current image, the reference samples are spaced apart from one another in the reference image by an interval corresponding to a result of the size comparison.