VR Server Stitched Image Transmission for Latency Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current virtual reality (VR) experience services face delays due to image processing and data transmission latency when users move abruptly in virtual spaces, necessitating a solution to reduce these delays and provide seamless VR experiences.
Innovation Solution
A method involving a server that generates and transmits stitched 360-degree images of virtual spaces based on user movements, using multiple virtual cameras with adjustable resolutions and operation states, and a user device that renders stereoscopic images based on head movements, minimizing latency by processing images locally.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the server processes and transmits 360-degree images in real-time to reflect user movements, then the VR experience responsiveness is improved, but the transmission latency increases
Solution Approach 1:
The patent segments the 360-degree image transmission into multiple lower-resolution image streams corresponding to different user movement states (e.g., walking, running, standing). Instead of transmitting a single high-resolution 360-degree image, the system transmits multiple segmented image sets that can be selectively delivered based on user state, reducing overall transmission data volume and latency while maintaining responsiveness.
Solution Approach 2:
The patent dynamically adjusts the image transmission strategy based on real-time user movement state detection. The server identifies whether the user is walking, running, or standing still, and adapts the resolution and number of transmitted images accordingly. This dynamic adaptation allows the system to optimize between transmission latency and responsiveness based on actual usage conditions.
2Measurement precision
If the server transmits high-resolution 360-degree images for all user movements, then the image quality is improved, but the transmission data volume increases
Solution Approach 1:
The patent applies different quality levels to different parts of the transmitted image data. Instead of uniformly transmitting high-resolution images for all scenarios, the system transmits high-resolution images only when necessary (e.g., when user is standing still for detailed observation), and uses lower-resolution images for dynamic movements (e.g., walking or running). This local quality differentiation optimizes the balance between image quality and data volume.
Solution Approach 2:
The patent changes the resolution parameter of transmitted images based on user movement state. When the user is detected to be walking or running, the system reduces the resolution parameter of transmitted images to decrease data volume. When the user is standing still, the system increases the resolution parameter to provide high-quality images. This parameter adaptation directly addresses the contradiction between image quality and data volume.
3Loss of time
If the user device performs local image rendering based on head movements, then the transmission latency is reduced, but the device processing requirement increases
Solution Approach 1:
The patent performs preliminary image rendering and preparation on the user device before actual use. The device pre-processes and stores image data locally, so when the user makes head movements, the system can quickly access and display pre-rendered images without requiring complex real-time rendering calculations. This preliminary action reduces the processing burden during actual interaction while maintaining low latency.
Data Source
AI summary
A user device and a server provide a VR experience service to a user without significant interruption in response to a user's movement. In particular, the server may create a stitched image of a 360-degree video of a predetermined virtual environment according to a user movement state and transmit the stitched image to a user device. The user device may i) receive, from a server through a wireless link, a stitched image of a 360-degree video of a predetermined virtual environment which is dynamically changed according to a user movement state and ii) render a stereoscopic image at a user's view-point which is dynamically changed according to a user's head movement, using the received stitched image of the 360-degree view of the predetermined virtual environment.