Foveated Rendering Reduces VR Data Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Virtual Reality (VR) devices face issues such as high cost, low resolution, and user discomfort due to Motion-to-Photon (MTP) redundancy, particularly in wireless environments, which hinder immersive experiences.
Innovation Solution
A stereoscopic image generating apparatus that reduces data transmission by utilizing an intra image of a right image as a basic inter image for a left image, synchronizing intra and inter pictures, and employing Foveated Rendering to increase resolution only where necessary, based on user head and body movements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full-resolution stereoscopic images are transmitted in wireless VR environments, then image quality is maintained, but data transmission volume and latency increase significantly
Solution Approach 1:
The patent applies Foveated Rendering technology to render images with high quality only in the foveal region (center of vision) and reduced quality in peripheral regions. This local differentiation of image quality reduces overall data transmission volume while maintaining perceived image quality, as the human eye is most sensitive to the central visual field.
Solution Approach 2:
The patent transmits only the necessary portion of image data by leveraging temporal redundancy between frames. By using inter-frame prediction and transmitting only motion-compensated differences rather than full frames, the system reduces data transmission volume while maintaining complete visual information where needed.
2Measurement precision
If high-resolution images are rendered for the entire field of view, then immersion is enhanced, but computational load and rendering time increase
Solution Approach 1:
The patent implements Foveated Rendering that dynamically adjusts rendering resolution based on the user's gaze direction. High-resolution rendering is applied only to the foveal region where the user is looking, while peripheral regions are rendered at lower resolution. This local quality differentiation reduces computational load and increases rendering speed while maintaining immersion in the focal area.
Solution Approach 2:
The rendering system dynamically adapts resolution based on real-time gaze tracking data. As the user moves their eyes or head, the high-resolution foveal region dynamically shifts position, and the rendering resolution adjusts accordingly. This dynamic adaptation maintains high productivity by always rendering high resolution only where the user is currently looking.
3Loss of time
If Motion-to-Photon latency is reduced to achieve immersive VR experience, then user comfort improves, but wireless data transmission constraints are exacerbated
Solution Approach 1:
The patent performs predictive rendering by anticipating future gaze positions based on current motion data. The system pre-renders images for predicted gaze locations rather than waiting for actual gaze changes, reducing the perceived Motion-to-Photon latency. This preliminary action allows the system to prepare high-priority visual data in advance, compensating for wireless transmission delays.
Solution Approach 2:
The patent dynamically changes rendering parameters including resolution, field of view, and texture quality based on motion state and gaze data. During high-motion states where latency is more critical, the system adjusts parameters to reduce rendering complexity and prioritize transmission of essential visual information, thereby reducing both rendering time and data volume while maintaining acceptable visual quality.
Data Source
AI summary
A stereoscopic image generating apparatus includes a FOV processor configured to receive head and body movements of a user mounted with a Head Mount Display (HMD) device and a Virtual Reality (VR) Vest device, a first eye image generating unit including an intra image independently generated for a first eye and an inter image derived from the intra image and configured to generate a redundant image up to a redundant angle of view greater by a specific criterion or more than a real angle of view which is currently displayed in a direction of the first eye from a center of the head direction of a user as a plurality of first eye images, and a second eye image generating unit configured to generate a plurality of second eye images based on the intra image among the plurality of first eye images for the second eye.