Inter-prediction Method for 360-Degree Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video compression techniques, such as H.264/AVC and HEVC, face challenges in efficiently encoding and decoding video data due to increasing data sizes, resolutions, and frame rates, necessitating improved encoding and decoding efficiency and quality.
Innovation Solution
A method for inter-prediction in video decoding that derives a motion vector from a bitstream, acquires reference samples from a corresponding region within a reference picture for external regions, and predicts blocks based on these samples, enhancing encoding efficiency and accuracy for 360-degree image processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video compression techniques (H.264/AVC, HEVC) are used, then existing encoding standards are maintained, but encoding efficiency and quality cannot keep up with increasing video size, resolution, and frame rate
Solution Approach 1:
The patent segments the reference picture into multiple regions (first reference block, second reference block, third reference block) and processes each region differently. External regions are mapped to corresponding regions within the reference picture, allowing efficient handling of different spatial zones with region-specific prediction strategies.
Solution Approach 2:
The patent introduces a new dimensional approach by mapping external regions (outside the reference picture boundary) to corresponding regions within the reference picture. This transforms the problem from 2D boundary handling to a wrapped dimensional space, enabling efficient 360-degree image processing.
2Measurement precision
If motion information is represented with full precision, then prediction accuracy is maintained, but the number of bits required increases
Solution Approach 1:
The patent applies different processing strategies to different regions. Internal regions use standard motion compensation, while external regions use region-mapping to corresponding internal regions. This local differentiation allows optimized bit allocation and reduced motion information requirements where applicable.
Solution Approach 2:
The patent copies motion information from corresponding regions within the reference picture to external regions. By wrapping external region motion vectors to their corresponding internal regions, the system reduces the bits needed to represent motion information while maintaining prediction accuracy for 360-degree content.
3Ease of manufacture
If reference samples from external regions are handled using conventional methods, then simple boundary processing is used, but prediction accuracy for 360-degree images deteriorates
Solution Approach 1:
Instead of extending the reference picture boundaries or using complex external region processing, the patent inverts the approach by mapping external regions back to corresponding regions within the reference picture. This inversion simplifies processing while improving 360-degree prediction accuracy.
Solution Approach 2:
The patent creates a universal region-mapping mechanism that handles both internal and external regions through a single correspondence relationship. The same mapping logic applies to all external regions, providing a unified approach that simplifies processing while maintaining high prediction accuracy for various 360-degree image types.
Data Source
AI summary
Disclosed are an inter-prediction method and an video decoding device. One embodiment of the present invention provides an inter-prediction method executed in an video decoding device, including deriving a motion vector of a current block based on motion information decoded from a bitstream; acquiring reference samples of a first reference block by using the motion vector, wherein reference samples of an external region located outside a reference picture among the first reference block are acquired from a corresponding region corresponding to the external region within the reference picture; and predicting the current block based on the acquired reference samples.