VR Image Processing Reducing Rendering Load via Region Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional VR image processing methods require significant computation to render stereoscopic images for both eyes, leading to time delays and low frame rates, causing vertiginous sensations in users.
Innovation Solution
A VR image processing method that reduces the amount of data to be rendered by determining a unique peripheral image view angle based on the left-eye and right-eye view angles, rendering viewpoint peripheral images at a lower resolution, and splicing these with viewpoint images to create complete images, while adjusting region sizes based on user displacement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If two groups of images are rendered simultaneously for left and right eyes, then stereoscopic effect is achieved, but computation time increases and frame rate decreases
Solution Approach 1:
The image rendering process is segmented into three distinct regions: viewpoint region (high resolution), transition region (medium resolution), and viewpoint peripheral region (low resolution). This segmentation allows different computational resources to be allocated to different regions, reducing overall computation time while preserving the critical stereoscopic effect in the viewpoint region.
Solution Approach 2:
Different resolution qualities are applied to different regions of the image. The viewpoint region maintains high resolution to preserve stereoscopic effect, the transition region uses medium resolution as a gradient, and the viewpoint peripheral region uses low resolution to reduce computation. This local quality differentiation resolves the contradiction by maintaining quality where needed while reducing overall computational load.
2Reliability
If two groups of images are rendered simultaneously for left and right eyes, then complete stereoscopic images are produced, but time delay increases
Solution Approach 1:
The image is divided into multiple regions that can be rendered at different speeds and resolutions. The viewpoint peripheral region requires minimal computation time, the transition region requires moderate time, and the viewpoint region requires maximum time but covers the smallest area. This segmentation significantly reduces total rendering time while preserving the stereoscopic effect in critical areas.
Solution Approach 2:
Instead of rendering complete high-resolution images for both eyes, the patent applies partial action by rendering only essential regions at high resolution (viewpoint region) and using lower resolution for peripheral regions. This partial rendering approach maintains the core stereoscopic effect while dramatically reducing computation time and time delay.
3Manufacturing precision
If high resolution is used for viewpoint peripheral regions, then image quality is maintained, but data rendering amount increases
Solution Approach 1:
The patent applies different quality levels to different regions: high resolution for the viewpoint region where stereoscopic effect is critical, medium resolution for the transition region, and low resolution for the viewpoint peripheral region. This local quality approach maintains necessary image quality in critical areas while reducing overall data rendering amount.
Solution Approach 2:
The patent uses partial action by rendering only the essential viewpoint region at high resolution and using lower resolution for the peripheral regions. This approach maintains adequate image quality for the stereoscopic effect while significantly reducing the total data rendering amount required.
4Manufacturing precision
If viewpoint region area is increased, then stereoscopic effect quality improves, but computation load increases
Solution Approach 1:
The patent concentrates high-quality rendering resources on the viewpoint region where stereoscopic effect quality is most critical, while using lower quality for the viewpoint peripheral region. This local quality differentiation improves stereoscopic effect quality in essential areas without proportionally increasing overall computation load.
Solution Approach 2:
The patent applies partial action by rendering only the essential viewpoint region at high quality and using lower quality for peripheral regions. This approach improves stereoscopic effect quality where it matters most while keeping the overall computation load manageable.
Data Source
AI summary
Provided are VR image processing method and apparatus. The method includes: rendering left-eye and right-eye viewpoint regions based on left-eye and right-eye view angles respectively, to obtain left-eye and right-eye viewpoint images; determining a candidate region based on positions of the left-eye and right-eye view angles, and selecting a point in the candidate region as a peripheral image view angle; rendering left-eye and right-eye viewpoint peripheral regions based on the peripheral image view angle, to obtain a same viewpoint peripheral image; splicing the viewpoint peripheral image with the left-eye viewpoint image and with the right-eye viewpoint image to obtain a left-eye complete image and a right-eye complete image; and reducing, when a displacement of a left-eye viewpoint or a right-eye viewpoint within a preset time period is less than a preset displacement, an area of a corresponding viewpoint region and increasing an area of a corresponding viewpoint peripheral region.