360-Degree Image Decoding with Projection-Based Inter Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems struggle with the massive data requirements of 360-degree images for virtual and augmented reality, necessitating improved performance in image encoding and decoding, particularly for 360-degree images.
Innovation Solution
A method for encoding and decoding 360-degree images that includes generating a predicted image using syntax information, combining it with a residual image, and reconstructing the image in various projection formats, such as Equi-Rectangular, CubeMap, and OctaHedron, while utilizing motion vector candidates and reference pictures for enhanced compression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If 360-degree images are captured and processed for virtual reality and augmented reality, then the realism and quality of media service are improved, but the amount of data generated increases massively and the performance of image processing systems becomes insufficient
Solution Approach 1:
The 360-degree image is divided into multiple projection faces (e.g., cube map faces, octahedron faces) that can be processed independently. Each face is encoded and decoded separately, allowing parallel processing and reducing the computational burden on the image processing system while maintaining overall image quality.
Solution Approach 2:
The patent transforms the 360-degree spherical image into multiple 2D projection formats (cube map, octahedron, icosahedron). This dimensional transformation allows standard 2D image processing techniques to be applied to 360-degree content, improving processing efficiency without sacrificing the immersive quality needed for VR/AR applications.
2Reliability
If image encoding and decoding is performed on 360-degree images, then the quality of virtual reality and augmented reality services is improved, but the processing time and computational resources increase
Solution Approach 1:
The patent performs image expansion and prediction operations using previously decoded reference pictures and motion vector candidates from prior frames. By preparing prediction data in advance and reusing it across multiple projection faces, the encoding and decoding process is accelerated without compromising image quality.
Solution Approach 2:
The patent uses motion compensation techniques where predicted blocks are copied from reference pictures based on motion vectors. This allows efficient reconstruction of image content by reusing existing data from previous frames, significantly reducing processing time while maintaining high image quality through accurate motion prediction.
3Productivity
If motion compensation and prediction are used for 360-degree image encoding, then compression performance is improved, but the complexity of the encoding process increases
Solution Approach 1:
The patent applies the same motion compensation and prediction mechanisms across all projection faces (cube map, octahedron, icosahedron) using unified syntax elements and processing logic. This universal approach improves compression performance through consistent motion modeling while reducing encoding complexity by avoiding face-specific customization of the encoding process.
Data Source
AI summary
A method of decoding an image, includes obtaining at least one offset for a picture, deriving a variable for scaling for the picture based on the at least one offset, and performing inter prediction based on the variable for scaling for the picture. The at least one offset is defined with a direction of scaling.