Video Encoder Motion Prediction with Selective Optical Flow
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in managing increased processing load and memory bandwidth during motion vector estimation and correction processes, particularly in next-generation video compression standards.
Innovation Solution
An encoder and decoder system that estimates motion vectors without using the current block's image, identifies a second region within a reference picture for correction, and permits the correction process only when the second region is entirely included in a first region, thereby reducing the need to read new reference images from the frame memory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the bi-directional optical flow process is always operated to correct the prediction image, then the prediction accuracy is improved, but the processing load and memory bandwidth increase
Solution Approach 1:
The patent applies partial action by selectively operating the bi-directional optical flow process only for blocks where the second region is included in the first region. This conditional application provides correction only when necessary (when regions overlap), rather than universally applying the correction process to all blocks, thus balancing prediction accuracy improvement with processing load reduction.
2Measurement precision
If the bi-directional optical flow process is always operated to correct the prediction image, then the prediction accuracy is improved, but the memory bandwidth increases
Solution Approach 1:
The patent reduces memory bandwidth consumption by applying the bi-directional optical flow correction process partially - only when the second region is included in the first region. This selective approach avoids unnecessary memory reads for blocks where correction is not needed, thereby optimizing memory bandwidth usage while maintaining prediction accuracy where it matters.
3Speed
If the correction process is permitted without region inclusion check, then the processing speed is improved, but the prediction accuracy deteriorates
Solution Approach 1:
The patent performs a preliminary check to determine whether the second region is included in the first region before deciding to apply the bi-directional optical flow correction. This preliminary region inclusion check enables the system to make an informed decision about whether correction is necessary, ensuring prediction accuracy is maintained when needed while avoiding unnecessary processing steps.
Data Source
AI summary
An encoding method is provided for encoding a picture to generate a coded stream. The encoding method includes: generating a first prediction image of a current block included in a current picture by referring to a first region included in a reference picture different from the current picture; operating a bi-directional optical flow process to generate a second prediction image based on the first prediction image by referring to a second region entirely included in the first region, the bi-directional optical flow process being performed without applying local illumination compensation; and encoding the current block based on the second prediction image.