Space Carving Near-Eye Display 3D Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Near-eye display systems face computational challenges in providing realistic augmented and virtual reality experiences due to the need for detailed 3D mapping of environments, which is resource-intensive and can be disrupted by obstacles like glass or mirrored surfaces.
Innovation Solution
The technology employs space carving based on user movement data, using sensors to identify navigable paths and store carved-out spaces in a 3D model, reducing computational complexity and enhancing reliability by assuming empty space where the user traverses, thereby improving the mapping process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If detailed 3D surface mesh mapping is used to provide realistic AR/VR experiences, then the realism and depth cue accuracy improve, but the computational cost and processing requirements increase significantly
Solution Approach 1:
The patent extracts only the essential navigable path information from the environment rather than creating a complete detailed 3D surface mesh. By taking out only the necessary data (paths users can walk through), the system achieves realistic AR/VR experiences without the full computational burden of complete environmental mapping.
Solution Approach 2:
The system performs partial mapping by focusing only on navigable paths rather than mapping the entire environment in detail. This partial action approach provides sufficient realism for AR/VR applications while significantly reducing computational requirements compared to complete detailed mapping.
2Measurement precision
If complete environmental mapping is performed to identify all objects and surfaces, then the accuracy of virtual object placement improves, but the processing time and computational resources increase
Solution Approach 1:
The patent extracts only the critical information needed for virtual object placement - the navigable paths - rather than performing complete environmental mapping. This extraction approach maintains sufficient accuracy for placing virtual objects realistically while dramatically reducing processing time.
Solution Approach 2:
The mapping process is segmented into essential components (navigable paths) rather than attempting to map every surface and object. This segmentation allows the system to achieve accurate virtual object placement on walkable surfaces without the time cost of complete environmental mapping.
3Manufacturing precision
If dense 3D surface mesh representation is created to enable realistic experiences, then the quality of depth cues and spatial accuracy improve, but the computational expense becomes prohibitive for NED systems
Solution Approach 1:
The system extracts only the essential spatial information - navigable paths through the environment - rather than creating dense 3D surface meshes. This extraction maintains sufficient spatial accuracy for realistic AR/VR experiences while making the computational expense acceptable for NED systems.
Solution Approach 2:
Instead of building up complete detailed 3D models and then extracting path information, the system inverts the approach by directly identifying navigable paths as the primary data structure. This inversion eliminates the need for computationally expensive dense mesh generation while maintaining necessary spatial accuracy.
Data Source
AI summary
Technology is described for (3D) space carving of a user environment based on movement through the user environment of one or more users wearing a near-eye display (NED) system. One or more sensors of the NED system provide sensor data from which a distance and direction of movement can be determined. Spatial dimensions for a navigable path can be represented based on user height data and user width data of the one or more users who have traversed the path. Space carving data identifying carved out space can be stored in a 3D space carving model of the user environment. The navigable paths can also be related to position data in another kind of 3D mapping like a 3D surface reconstruction mesh model of the user environment generated from depth images.