360 Image Projection Using Multi-View 2D Blocks and Blending
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods for 360-degree images, such as equirectangular projections, suffer from distortion and edge cutoff issues, making it difficult to analyze and process these images effectively.
Innovation Solution
An electronic apparatus and method that converts a spherical image into multiple 2D images by adjusting the projection viewpoint and direction, minimizing distortion, and using region-specific filters to enhance image analysis and processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a still image is divided into multiple blocks for processing, then processing efficiency is improved, but block boundaries become visible and degrade image quality
Solution Approach 1:
The patent applies asymmetric overlap between adjacent blocks, where the overlap region extends differently in horizontal and vertical directions. This asymmetric design allows the blending operation to smoothly transition across block boundaries while maintaining processing efficiency, effectively hiding the block boundaries that would otherwise degrade image quality.
Solution Approach 2:
The patent introduces a blending operation as an intermediary process between block processing and final image composition. This blending operation acts as a mediator that smooths transitions at block boundaries by combining pixel values from overlapping regions, thereby eliminating visible block boundaries while preserving the benefits of divided processing.
2Manufacturing precision
If block size is increased to reduce block boundaries, then image quality is improved, but processing efficiency decreases
Solution Approach 1:
The patent employs dynamic block processing where the effective block size varies depending on the presence of overlapping regions. By dynamically adjusting the processing scope through overlapping blocks with different dimensions, the system achieves both fine-grained processing for efficiency and smooth transitions for quality, avoiding the need for uniformly large blocks.
Solution Approach 2:
The patent resolves the contradiction by transitioning from a single-dimension block size parameter to a two-dimensional block structure with both width and height dimensions, where one dimension may be larger than the other. This dimensional change allows the system to optimize for both processing efficiency and image quality simultaneously by creating rectangular blocks with strategic overlap patterns.
3Manufacturing precision
If overlapping regions are added between blocks, then block boundaries are hidden, but processing complexity increases
Solution Approach 1:
The patent segments the image processing into distinct phases: block division, overlapping region identification, and blending operation. This segmentation of the processing workflow allows each stage to be handled independently with clear algorithms, reducing overall processing complexity while achieving the goal of hiding block boundaries through systematic overlap management.
Solution Approach 2:
The patent manages processing complexity by changing key parameters such as overlap width and height, and blending weights in a systematic manner. By establishing clear parameter relationships and using consistent blending formulas across all overlapping regions, the system achieves smooth transitions without requiring complex adaptive processing for each region.
Data Source
Figure 1~2b
Figure 2c~3b
Figure 3c~4a
AI summary
An electronic apparatus is provided. The apparatus includes a storage and a processor configured to convert an original image to a spherical image, to acquire a plurality of two-dimensional (2D) images corresponding to a respective plurality of points on the spherical image by projecting the spherical image onto a plane with reference to the respective plurality of points on the spherical image, and to store the acquired 2D images in the storage.