Optical Flow Video Encoding With Hyperprior-Based Bitrate Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding technologies face challenges in managing bitrate overheads, particularly in high-definition and ultra-high-definition video transmission, due to the lack of efficient methods for reducing redundancy between video frames.
Innovation Solution
An encoding method that utilizes motion estimation to determine optical flow hyperprior information and residual information based on prior and spatial domain information, reducing bitrate by encoding these components separately to minimize data redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional video compression technologies (H265, H266) are used, then bandwidth and storage requirements increase for high definition videos, but video quality requirements cannot be met
Solution Approach 1:
The patent segments the video encoding process into multiple components: motion estimation to obtain optical flow, hyperprior information determination, probability distribution estimation, and entropy encoding. By dividing the encoding task into these segments, the system achieves better compression efficiency and video quality while controlling bandwidth and storage requirements
Solution Approach 2:
The patent performs preliminary actions by determining hyperprior information about optical flow before actual encoding. The system estimates probability distributions in advance and uses this prior knowledge to guide the encoding process, which improves compression efficiency and reduces the bandwidth and storage needed for high definition videos
2Productivity
If inter encoding is performed on all video frames to reduce bitrate overheads, then compression efficiency improves, but complexity of probability distribution estimation increases
Solution Approach 1:
The patent applies local quality by performing probability distribution estimation selectively rather than uniformly across all frames. The system identifies frames that benefit most from inter encoding and applies the complex probability estimation only where needed, while using simpler methods for other frames, thus improving compression efficiency without uniformly increasing complexity
Solution Approach 2:
The patent changes parameters by using hyperprior information to adapt the probability distribution estimation process. Instead of using fixed complex models for all frames, the system adjusts the estimation parameters based on the determined hyperprior information about optical flow, which simplifies the overall process while maintaining compression efficiency
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
This application provides an encoding method, a decoding method, and an electronic device. The method includes: performing motion estimation on a current frame, to obtain a first optical flow; obtaining first optical flow hyperprior information based on feature information of the first optical flow and prior information of the first optical flow; encoding the first optical flow hyperprior information, to obtain a first bitstream; performing probability estimation based on second optical flow hyperprior information and prior information of the first optical flow, to obtain an optical flow probability distribution; performing entropy encoding on the feature information of the first optical flow based on the optical flow probability distribution, to obtain a second bitstream; performing inter prediction on the current frame, to obtain prediction information; determining residual information based on the prediction information and an original picture; and encoding the residual information, to obtain a third bitstream. Compared with the conventional technology in which determined first optical flow hyperprior information is used to determine an optical flow probability distribution of all samples in the current frame, in this application in which the first optical flow hyperprior information is used to determine an optical flow probability distribution of a part of samples in the current frame, bitrate overheads can be reduced.