Adaptive Interpolation Offset for Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression technologies, such as H.264/AVC, fail to effectively predict motion in images with brightness changes over time, like lighting changes, due to the use of fixed filters for interpolating reference frames, which limits compression efficiency.
Innovation Solution
An apparatus and method for encoding images that interpolates a reference frame by adding an offset to the interpolated pixels, allowing for more accurate motion prediction by assimilating the reference frame into the current frame, and encodes residual signals with flexible transform and quantization schemes based on specific encoding schemes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a fixed filter is used to interpolate the reference frame, then the interpolation process is simple and fast, but the motion prediction accuracy deteriorates when brightness changes occur
Solution Approach 1:
The patent applies the Dynamics principle by making the interpolation filter adaptive rather than fixed. The filter coefficients are dynamically adjusted based on the gradient information calculated from the reference frame pixels. This allows the filtering process to adapt to local brightness changes and edge characteristics, improving motion prediction accuracy while maintaining reasonable computational complexity.
Solution Approach 2:
The patent implements Parameter changes by modifying the filter coefficients based on gradient calculations. Instead of using fixed coefficients, the filter parameters are changed according to the local image characteristics (gradient magnitude and direction). This enables the interpolation to account for brightness changes and improve prediction accuracy in regions with varying illumination.
2Loss of information
If the reference frame is interpolated with fractional pixel precision, then compression efficiency is improved, but the complexity of the encoding process increases
Solution Approach 1:
The patent applies the Preliminary action principle by performing gradient calculations and determining optimal filter coefficients before the actual interpolation process. This preliminary preparation allows the subsequent interpolation to proceed efficiently with pre-determined parameters, reducing the overall computational burden despite the increased precision requirements.
Solution Approach 2:
The patent implements Segmentation by dividing the interpolation process into distinct stages: gradient calculation, filter coefficient determination, and actual interpolation. This segmentation allows each stage to be optimized independently and enables selective application of complex processing only where necessary (e.g., only in regions with significant gradient variations).
Data Source
AI summary
A method for encoding pixels in an image, includes: encoding an offset to be applied to a reference frame; generating predicted pixels of the pixels in the image, based on interpolating pixels in the reference frame and then adding the offset to the interpolated pixels; and encoding residual signals that are differences between the pixels in the image and the predicted pixels, wherein the encoding of the residual signals comprises: skipping both transform and quantization with respect to residual signals of which an encoding scheme corresponds to skipping of both transform and quantization, the encoding scheme related to transform and quantization, skipping transform and performing quantization with respect to residual signals of which the encoding scheme corresponds to skipping of transform, and performing transform and quantization with respect to residual signals of which the encoding scheme corresponds to skipping neither transform nor quantization.