Video Coding with Optical Flow for Controlled Datamoshing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video coding techniques suffer from low coding efficiency and lack structured control over glitch artifacts, particularly in datamoshing effects, as they do not effectively track real object movements during motion compensation.
Innovation Solution
Implement an artifact coding tool that enables optical flow-based motion compensation and defines a region of interest (ROI) for selective coding tools, ensuring consistent motion tracking and controlled glitch artifacts by disabling certain tools like in-loop filters and allowing only uni-prediction within the ROI.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional video coding techniques are used, then coding efficiency is maintained at standard levels, but the ability to generate controlled glitch artifacts and track real object motion is insufficient
Solution Approach 1:
The patent changes the motion compensation parameters by switching from conventional block-based motion estimation to optical flow-based motion estimation. This parameter change enables pixel-level motion tracking which reliably generates datamoshing artifacts while maintaining coding efficiency through selective application of coding tools in ROI and non-ROI regions.
Solution Approach 2:
The patent applies different coding tool configurations to different regions of the video block. Within the region of interest (ROI), certain coding tools are disabled to preserve motion artifacts, while in non-ROI regions, standard coding tools are applied. This local differentiation enables reliable artifact generation in specific areas without sacrificing overall coding efficiency.
2Measurement precision
If optical flow-based motion compensation is implemented, then real object motion tracking is improved, but device complexity increases
Solution Approach 1:
The patent segments the video block into two distinct regions: ROI (region of interest) and non-ROI regions. This segmentation allows optical flow-based motion compensation to be applied selectively only where needed for artifact generation, reducing the overall computational complexity while maintaining high motion tracking precision in critical areas.
Solution Approach 2:
The patent applies optical flow-based motion compensation and selective coding tool disabling only partially - specifically within the ROI regions where artifact generation is desired. In non-ROI regions, conventional coding methods are used. This partial application reduces device complexity compared to applying optical flow throughout the entire video block, while still achieving the required motion tracking precision for artifact generation.
Data Source
AI summary
Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: determining, during a conversion between a target video block of a video and a bitstream of the video, whether an artifact coding tool is enabled, the artifact coding tool being used to generate an artifact effect in the video; in response to the artifact coding tool being enabled, determining at least one candidate coding tool from a plurality of coding tools; and performing the conversion by enabling the at least one candidate coding tool.