Inter-View Motion Merge Candidate Derivation in 3D Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In 3D-HEVC, when motion is uniform, the inter-view motion merge process cannot derive motion information for the current block from the reference block, rendering the inter-view motion merge process unusable.
Innovation Solution
A method and device for generating an inter-view motion merge candidate by determining if the inter-view motion merge process is possible for the current block based on the encoding information of the inter-view reference block. If not possible, the method generates a motion merge candidate using encoding information from a neighboring block that is spatially adjacent to the inter-view reference block.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the inter-view motion merge process uses the reference block in 3D-HEVC, then the motion information can be derived for inter-coded blocks, but the process becomes unusable when the reference block is intra-coded or motion is uniform
Solution Approach 1:
The patent introduces an intermediary mechanism that checks the coding mode of the reference block and selectively applies different motion merge strategies. When the reference block is intra-coded, the system uses spatial neighboring blocks as intermediaries to provide motion information, thereby maintaining the applicability of the inter-view motion merge process across different block types
Solution Approach 2:
The patent dynamically changes the source of motion information based on the coding mode parameter of the reference block. When the reference block is inter-coded, motion information is derived from the reference block; when intra-coded, the system switches to using spatial neighboring blocks, thus adapting to different parameter states to maintain encoding efficiency
2Reliability
If the inter-view motion merge process is restricted to only inter-coded reference blocks, then the motion information derivation is reliable, but the process complexity increases and encoding efficiency decreases
Solution Approach 1:
The patent segments the motion merge process into two distinct paths: one for inter-coded reference blocks and another for intra-coded reference blocks. This segmentation allows each path to be optimized independently, maintaining reliability for inter-coded blocks while providing a fallback mechanism for intra-coded blocks, thereby reducing overall process complexity
3Adaptability or versatility
If spatial neighboring blocks are used to derive motion merge candidates, then the inter-view motion merge process can be applied to intra-coded reference blocks, but additional computation and memory access are required
Solution Approach 1:
The patent applies partial action by only using spatial neighboring blocks when absolutely necessary (i.e., when the reference block is intra-coded). For the more common case of inter-coded reference blocks, the original efficient inter-view motion merge process is used, thus avoiding unnecessary computation and memory access while maintaining adaptability
Data Source
AI summary
The present invention provides a method and a device for deriving an inter-view motion merging candidate. A method for deriving an inter-view motion merging candidate, according to an embodiment of the present invention, can comprise the steps of: on the basis of encoding information of an inter-view reference block derived by means of a variation vector of a current block, determining whether or not inter-view motion merging of the current block is possible; and, if inter-view motion merging of the current block is not possible, generating an inter-view motion merging candidate of the current block by using encoding information of an adjacent block that is spatially adjacent to the inter-view reference block.