3D Video Encoding Reusing Motion Vectors Across Views
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for encoding and decoding 3D video lack efficiency in transferring and storing high-resolution video data, as they do not effectively reuse motion information across different views, leading to increased data transmission and processing complexity.
Innovation Solution
A method and device that reuse motion information, such as motion vectors, between different views in 3D video encoding and decoding, by determining and transmitting motion information of a current block using motion information from a reference block, including the use of global disparity vectors and quad tree structures to optimize data transmission and processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion information is independently determined for each block in 3D video encoding, then decoding accuracy is improved, but data transmission volume and processing complexity increase
Solution Approach 1:
The patent copies motion information from reference blocks in other views to current blocks, using inter-view prediction to reuse motion vectors and disparity vectors across multiple views. This copying approach maintains decoding accuracy while significantly reducing the amount of motion information that needs to be transmitted for each block independently.
Solution Approach 2:
The patent makes motion information universal across different views by establishing that motion vectors and disparity vectors can be shared and reused across multiple views. A single motion information set serves multiple blocks across different views, reducing redundancy and transmission requirements while maintaining accuracy through adaptive selection.
2Measurement precision
If motion information is independently determined for each block in 3D video encoding, then decoding accuracy is improved, but device complexity increases
Solution Approach 1:
By copying motion information from reference blocks rather than calculating it independently for each block, the patent reduces the computational burden on encoding and decoding devices. The copying operation is computationally simpler than independent motion estimation, thereby reducing device complexity while maintaining adequate decoding accuracy through selective usage.
Solution Approach 2:
The patent applies partial action by selectively using inter-view prediction for certain blocks rather than all blocks. The system determines whether to use copied motion information based on conditions such as view correlation and block characteristics, applying the simplified approach only where beneficial, thus balancing accuracy requirements with complexity reduction.
3Measurement precision
If high-resolution video data is transferred using existing media, then video quality is maintained, but transfer cost increases
Solution Approach 1:
The patent copies and reuses motion information across multiple views and blocks, which directly reduces the amount of data that needs to be transmitted. This copying mechanism maintains video quality by preserving essential motion characteristics while significantly reducing the bandwidth and storage requirements, thereby lowering transfer and storage costs.
Solution Approach 2:
The patent discards redundant motion information that would otherwise need to be transmitted for each block independently, and recovers the necessary motion data through inter-view prediction from reference blocks. This selective discarding and recovering approach reduces transmission volume while maintaining the quality needed for accurate decoding.
4Measurement precision
If high-resolution video data is stored in existing storage media, then video quality is maintained, but storage cost increases
Solution Approach 1:
By copying motion information across views and blocks, the patent reduces the total volume of video data that needs to be stored. The redundant motion information is eliminated through inter-view prediction, allowing high-resolution video to be stored in existing media with reduced storage costs while maintaining video quality through efficient compression.
Data Source
AI summary
A video decoding method that includes: receiving information for deriving motion information of a current block; deriving the motion information of the current block based on the received information for deriving the motion information; and performing prediction to generate predicted pixels of the current block based on the motion information of the current block, wherein the motion information of the current block is determined by using motion information of a reference block, wherein the reference block is determined based on a specific disparity vector, wherein the specific disparity vector is determined for an area in a picture to which the current block belongs, wherein the area which the specific disparity vector is determined is split based on a quad tree structure, and wherein the current block is a block of a texture picture and the reference block is a block in a reference view is disclosed.