360-Degree Video Block Prediction Using Region-Based Spatial Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing compression methods for 360-degree videos do not effectively utilize spatial correlations between different viewpoints, leading to inefficiencies in encoding and decoding processes.
Innovation Solution
A region-based inter/intra prediction technique is employed, determining spatial correlation degrees between blocks in 360-degree video images to enhance encoding/decoding efficiency by selecting optimal candidate blocks for prediction, including using cube map and equi-rectangular projections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing compression methods are applied to 360-degree videos, then compatibility with standard video coding is maintained, but encoding efficiency is insufficient due to lack of spatial correlation utilization
Solution Approach 1:
The patent applies different prediction techniques to different regions (blocks) of the 360-degree video based on their spatial correlation characteristics. Each block is analyzed to determine its correlation with neighboring blocks, and appropriate prediction methods are selectively applied to optimize encoding efficiency for each specific region while maintaining overall system versatility.
Solution Approach 2:
The patent dynamically adjusts the prediction process by determining spatial correlation degrees between blocks and adapting the prediction technique accordingly. The system can switch between inter prediction, intra prediction, or skip modes based on real-time analysis of spatial correlations, enabling efficient adaptation to varying video content characteristics.
2Productivity
If region-based prediction is implemented, then encoding efficiency is improved through optimal block selection, but device complexity increases due to additional correlation analysis
Solution Approach 1:
The patent performs preliminary analysis of spatial correlation degrees between blocks before the actual encoding process. By pre-determining which blocks have high spatial correlation with their neighbors, the system can efficiently select prediction techniques during encoding without performing complex correlation analysis in real-time, thus reducing overall computational complexity.
Solution Approach 2:
The prediction system utilizes already-encoded neighboring blocks as references for predicting current blocks. The encoded data from neighboring blocks serves the dual purpose of both encoding those blocks and providing prediction information for adjacent blocks, making the system self-sufficient and reducing the need for separate correlation analysis mechanisms.
3Measurement precision
If multiple prediction candidates are evaluated, then prediction accuracy is improved, but computational resources are consumed
Solution Approach 1:
Instead of evaluating all possible prediction candidates exhaustively, the patent applies partial action by considering only the most promising candidates based on pre-determined spatial correlation degrees. The system evaluates a limited set of high-probability prediction candidates rather than all possibilities, achieving sufficient prediction accuracy while significantly reducing computational resource consumption.
Solution Approach 2:
The patent changes the parameter of prediction candidate selection based on spatial correlation degree thresholds. By adjusting which candidates are evaluated based on their correlation strength with neighboring blocks, the system dynamically balances prediction accuracy requirements against computational resource availability, optimizing the trade-off for different video conditions.
Data Source
AI summary
Disclosed are region-based image encoding/decoding method and apparatus. The method of decoding an image includes: determining a projection format of a 360-degree video image; determining, on the basis of the determined projection format, a spatial correlation degree between a current block and a neighboring block; and determining, on the basis of the determined spatial correlation degree, a candidate block for predicting the current block.