Multi-Plane Image Conversion with Empty-Space Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for converting multi-plane 3D data into a 2D atlas image result in a significantly increased image size and reduced compression efficiency due to the presence of numerous empty spaces in the 2D atlas image.
Innovation Solution
The method involves generating a multi-plane image reconfigured into layers based on multi-view image data, aggregating these layers into aggregated layers, and converting the multi-plane image into a 2D atlas image, using techniques like HEVC and VVC for compression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If patch regions from each layer image are gathered into one 2D atlas image using conventional conversion scheme, then the multi-plane image can be converted into a 2D atlas image, but the size of the 2D atlas image is greatly increased and compression efficiency deteriorates
Solution Approach 1:
The patent merges multiple layer images into a single aggregated layer by combining patch regions from different depth layers. This consolidation eliminates the need for separate layer storage and reduces the overall atlas image size while maintaining all necessary visual information from the multi-plane image.
Solution Approach 2:
The patent reorganizes multi-dimensional layer data into a two-dimensional atlas structure by projecting patch regions from multiple depth layers onto a single 2D plane. This dimensional transformation allows efficient storage and retrieval while reducing the quantity of data required compared to storing each layer separately.
2Ease of manufacture
If patch regions from each layer image are gathered into one 2D atlas image using conventional conversion scheme, then the multi-plane image can be converted into a 2D atlas image, but a larger number of empty spaces occur in the 2D atlas image
Solution Approach 1:
The patent segments the aggregated layer into multiple sub-regions corresponding to different depth layers, allowing selective rendering of only necessary patch regions. This segmentation enables the system to avoid rendering empty spaces by only processing and displaying regions that contain actual patch data from the multi-plane image.
Solution Approach 2:
The patent implements dynamic patch region rendering where the system selectively activates and renders only the patch regions that are currently needed based on the viewing angle and depth information. This dynamic approach minimizes the display of empty spaces by adapting the rendered content to the specific viewing conditions.
Data Source
AI summary
Disclosed herein are an image compression apparatus and method. The image conversion method includes generating a multi-plane image reconfigured into layers in a depth direction based on multiple pieces of multi-view image data, generating an aggregated layer by aggregating the layers of the multi-plane image into at least one layer, and converting a multi-plane image including the aggregated layer into a two-dimensional (2D) atlas image.