Remote Driving Assistance Using VR Scene Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In remote assisting automatic driving, users rely on planar video images to perceive the distance between unmanned devices and obstacles, leading to inadequate environment engagement and perception, potentially resulting in incorrect instructions and safety threats.
Innovation Solution
The method involves capturing and processing scene information from multiple environment angles to generate a Virtual Reality (VR) scene, which is displayed to the user, enabling accurate remote assisting driving operations by providing a more immersive and accurate representation of the environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If planar video images are used for remote assisting driving, then the system complexity is low, but the user's environment perception accuracy deteriorates
Solution Approach 1:
The patent transitions from 2D planar video images to 3D VR (Virtual Reality) scenes, adding a spatial dimension to the visual information. This dimensional change enables users to perceive depth, distance, and spatial relationships more accurately, directly addressing the limitation of planar images in providing environment perception accuracy while maintaining manageable system complexity through standardized VR processing pipelines.
Solution Approach 2:
The patent introduces a VR scene generation module as an intermediary between the unmanned device's camera system and the user's display interface. This intermediary processes raw video feeds from multiple angles, synthesizes them into coherent 3D VR environments, and presents them to users, thereby enhancing perception accuracy without requiring direct complex hardware modifications to the unmanned device itself.
2Measurement precision
If multiple environment angles are captured and processed into VR scene, then the environment perception accuracy is improved, but the device complexity increases
Solution Approach 1:
The patent divides the environment capture task into multiple discrete camera angles (e.g., front, rear, left, right views) captured simultaneously by separate imaging units. Each angle is processed independently into its own video stream, which are then synthesized into a comprehensive VR scene. This segmentation allows the system to achieve accurate multi-angle perception while managing complexity through modular, independent processing channels.
Solution Approach 2:
The patent creates virtual copies of the physical environment from multiple camera perspectives, synthesizing these copies into a unified VR representation. Instead of requiring a single complex omnidirectional camera, the system uses multiple standard cameras whose feeds are digitally replicated and repositioned to construct the VR scene, thereby achieving accurate environment perception using conventional, manageable components.
3Measurement precision
If VR scene is displayed to user, then the remote assisting driving accuracy is improved, but the information processing time increases
Solution Approach 1:
The patent performs preliminary processing of video feeds from multiple angles simultaneously as they are captured, pre-synchronizing and pre-aligning the streams before VR scene synthesis. This preliminary action ensures that when the VR scene is generated and displayed, the processing is already optimized, minimizing additional latency and reducing the overall information processing time while maintaining high remote assisting driving accuracy.
Data Source
AI summary
The present disclosure provides a method, a device for assisting driving, an unmanned device and a readable storage medium, where the method includes: the unmanned device triggers remote assisting automatic driving function, receiving the scene information captured by the unmanned device; processing the scene information, generating a VR scene; displaying the VR scene for a user, so that the user can trigger a remote assisting driving operation according to the VR scene; where the remote assisting driving operation is used to control the unmanned device to execute a driving operation. Thus providing a VR scene of the environment where the unmanned device is located, improving effectively user's environment engagement and perception of actual environment, which is helpful for the user to trigger an accurate remote assisting driving instruction, thus to ensure the driving safety of the unmanned device.