Cubic Face Image Assembly for 360-Degree Panorama Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image and video processing techniques for 360-degree panoramic videos face inefficiencies in storage and transmission due to the large data requirements of six cubic face images, which are not optimally formatted for spatial and temporal redundancy exploitation.
Innovation Solution
The method involves converting sets of six cubic faces into rectangular assembled images by maximizing continuous boundaries and minimizing discontinuous boundaries, either through unfolding and assembling into connected images or filling gaps with padding data to create fully-connected images, allowing for efficient video coding and compression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If six cubic face images are used to represent 360-degree panoramic video, then the panoramic view coverage is improved, but the storage space and transmission bandwidth requirements increase significantly
Solution Approach 1:
The patent combines six separate cubic face images into a single rectangular assembled image by arranging them in a 2×3 grid configuration. This merging reduces the number of independent image files from six to one, simplifying storage and transmission while preserving the complete 360-degree panoramic content.
Solution Approach 2:
The patent transforms the six separate 2D cubic face images into a single rectangular 2D assembled image by changing their spatial arrangement from a 3D cube projection to a flattened 2D grid layout. This dimensional transformation enables more efficient compression and processing while maintaining all panoramic view information.
2Device complexity
If conventional video coding is applied directly to six separate cubic face images, then the processing simplicity is maintained, but the compression efficiency is suboptimal due to inability to exploit spatial and temporal redundancy across faces
Solution Approach 1:
The patent merges six separate cubic face images into a single rectangular assembled image, enabling conventional video coding tools to exploit spatial and temporal redundancy across all faces simultaneously. This unified structure allows motion compensation and prediction to operate across face boundaries, significantly improving compression efficiency.
Solution Approach 2:
The patent segments the rectangular assembled image into six virtual cubic face regions during the coding process, allowing the encoder to treat it as a unified structure for redundancy exploitation while maintaining the logical separation of individual faces for proper rendering and display.
3Adaptability or versatility
If the six cubic faces are assembled into rectangular images with discontinuous boundaries, then the assembly flexibility is maintained, but the spatial redundancy exploitation is reduced due to broken continuity across face edges
Solution Approach 1:
The patent dynamically adjusts the assembly configuration of the six cubic faces into rectangular images, optimizing the arrangement to maximize continuous boundaries. The system can adaptively choose different assembly patterns (such as 2×3 grid) based on content characteristics to maintain spatial continuity while preserving assembly flexibility.
Solution Approach 2:
The patent employs asymmetric assembly arrangements of the six cubic faces in the rectangular image, strategically positioning faces to maximize the length and continuity of shared boundaries. This asymmetric optimization ensures that adjacent faces in the rectangular assembly share continuous spatial information, enhancing redundancy exploitation.
Data Source
AI summary
Methods and apparatus of processing cube face images are disclosed. According one method, each set of six cubic faces is converted into one rectangular assembled image by assembling each set of six cubic faces to maximize a number of continuous boundaries and to minimize a number of discontinuous boundaries. Each continuous boundary corresponds to one boundary between two connected faces with continuous contents from one face to another face. Each discontinuous boundary corresponds to one boundary between two connected faces with discontinuous contents from one face to another face. The method may further comprise applying video coding to the video sequence outputting the compressed data of the video sequence. According to another method, a fully-connected cubic-face image representing an unfolded image from the six faces of the cube is generated and the blank areas are filled with padding data to form a rectangular assembled image.