360-Degree Image Decoding With Recursive Block Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems struggle with the massive data generated for 360-degree images in virtual and augmented reality, requiring improved performance in encoding and decoding methods.
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, OctaHedron, and IcoSahedral, to enhance compression performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If multi-view images are captured with multiple cameras for 360-degree images, then the realism and quality of virtual reality and augmented reality services are improved, but the amount of data generated increases massively
Solution Approach 1:
The patent divides the 360-degree image into multiple projection formats (Equi-Rectangular, CubeMap, OctaHedron, IcoSahedral) and processes each format separately with optimized encoding parameters. This segmentation allows the system to handle different regions of the image with appropriate compression strategies, reducing overall data volume while maintaining quality.
Solution Approach 2:
The patent applies different encoding quality settings to different regions of the 360-degree image based on their importance. Regions with higher visual importance retain more detail, while less critical regions use higher compression. This local quality approach maintains perceived image quality while significantly reducing total data amount.
2Manufacturing precision
If high-resolution and high-quality images are processed, then the quality of virtual reality and augmented reality services is improved, but the performance of the image processing system becomes insufficient
Solution Approach 1:
The patent implements dynamic encoding parameters that adapt to the specific characteristics of each projection format and image region. The system dynamically adjusts compression ratios, block sizes, and transformation methods based on local image content and format requirements, optimizing processing performance for high-quality images without uniform quality loss.
Solution Approach 2:
The patent changes multiple encoding parameters including transformation type, quantization matrices, and prediction modes based on the projection format and image characteristics. These parameter changes enable efficient processing of high-resolution images by selecting optimal encoding settings for each specific case, improving overall system productivity.
3Device complexity
If conventional image encoding methods are used for 360-degree images, then the encoding process is simple, but the compression performance is insufficient for large data volumes
Solution Approach 1:
The patent creates a universal encoding framework that handles multiple projection formats (Equi-Rectangular, CubeMap, OctaHedron, IcoSahedral) through a single integrated process. This multi-functional approach maintains relatively simple encoding complexity by using consistent processing steps across different formats while achieving superior compression performance through format-specific optimizations.
Data Source
AI summary
Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.