Specific Operation Prediction for Rotated Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video compression technologies face challenges in efficiently handling videos with rotational and zooming operations, such as those captured from mobile devices or shared screens, which can result in distorted images, leading to increased complexity and inefficiency in compression and transmission.
Innovation Solution
The implementation of specific-operation prediction (SOP) methods and apparatuses that enable decoding and encoding of video bitstreams by determining and applying rotation and zoom operations to reference pictures, allowing for improved prediction and compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video compression is used on videos with rotation and zoom operations, then the videos can be compressed and transmitted, but the distorted images lead to increased complexity and reduced compression efficiency
Solution Approach 1:
The patent applies specific operations (rotation and zoom) to reference pictures before performing motion compensation prediction. This preliminary action transforms the reference pictures to match the current picture's orientation and scale, enabling more accurate motion estimation and reducing the residual error, thereby improving compression efficiency without significantly increasing complexity
Solution Approach 2:
The patent introduces specific operation parameters (rotation angle, zoom scale) to transform reference pictures. By changing the parameters of the reference pictures through rotation and zoom operations, the system adapts to videos captured with mobile devices or screen sharing operations, improving prediction accuracy and compression performance
2Productivity
If specific operation prediction is applied to handle rotated and zoomed images, then compression efficiency is improved, but additional processing steps increase device complexity
Solution Approach 1:
The encoder performs rotation and zoom operations on reference pictures before motion compensation, preparing transformed reference pictures in advance. This preliminary transformation enables more accurate prediction for rotated and zoomed video content, improving communication efficiency while keeping the additional complexity manageable through reusable transformed references
Solution Approach 2:
The patent creates transformed copies of reference pictures through rotation and zoom operations. These copied and transformed reference pictures are then used for motion compensation prediction, allowing the system to handle rotated and zoomed content efficiently without processing the entire original reference picture repeatedly
Data Source
AI summary
A decoding method includes: decoding from the video bitstream a prediction status parameter indicative of whether SOP is enabled; if the SOP is enabled, determining a SOP type indicative of a specific operation for predicting a current picture of the video bitstream; and determining a predicted picture by applying the specific operation to at least one of a reference picture for the current picture and the current picture, in which the specific operation includes at least one of rotation and zoom. An encoding method includes: if SOP is enabled, determining SOP data indicative of a specific operation for predicting a current picture of a video sequence; determining a predicted picture by applying the specific operation to at least one of a reference picture for the current picture and the current picture, in which the specific operation includes at least one of rotation and zoom; and encoding the SOP data.