Depth Video Coding Using Endurable View Synthesis Distortion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multiview Video Coding (MVC) requires efficient compression techniques to reduce bandwidth requirements due to high levels of statistical dependencies between and within groups of pictures from multiple camera views, leading to large data quantities.
Innovation Solution
The implementation of a system that uses endurable view synthesis distortion (EVSD) for depth video coding, incorporating a distortion estimator component to generate distortion values based on image capturing device parameters, which are then used by an encoder component for rate-distortion optimization and bit allocation in multiview video coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional multiview video coding is used to capture scenes from multiple viewpoints, then view synthesis quality is improved, but bandwidth requirements increase due to large data quantities
Solution Approach 1:
The patent extracts only the essential depth information from multiview video sequences and transmits it separately from texture data. By separating depth map transmission from full video data transmission, the system achieves accurate view synthesis with significantly reduced bandwidth requirements, as depth maps require much less data than complete video frames
Solution Approach 2:
The patent introduces depth maps as an intermediary representation that mediates between multiple camera views and the synthesized output. These depth maps serve as compact intermediaries that enable efficient view synthesis by providing geometric information without transmitting all the detailed visual data from each viewpoint
2Productivity
If depth video coding with view synthesis distortion estimation is implemented, then coding efficiency is improved, but system complexity increases due to additional distortion estimation components
Solution Approach 1:
The patent performs view synthesis distortion estimation during the encoding phase rather than during decoding or rendering. By calculating distortion metrics in advance during compression, the system optimizes bit allocation and coding decisions beforehand, improving overall coding efficiency while keeping the decoding process relatively simple
Solution Approach 2:
The patent implements a feedback mechanism where distortion estimation results from view synthesis are fed back into the coding process. This feedback loop allows the encoder to adjust quantization parameters and bit allocation based on actual synthesized view quality, continuously optimizing coding efficiency without requiring complex real-time adjustments at the decoder
Data Source
AI summary
The disclosed subject matter provides depth video coding using endurable view synthesis distortion (EVSD). In particular, a distortion estimator component receives at least one parameter associated with at least one image capturing device and generates a distortion value based on the at least one parameter. An encoder component encodes a multiview input stream based at least in part on the distortion value. As such, compression of depth information can provide for reduced bandwidth consumption for dissemination of encoded multiview content for applications such as 3D video, free viewpoint TV, etc.