AR Virtual Object Placement With Occlusion-Aware Surface Anchoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for displaying and manipulating virtual objects in augmented reality environments are cumbersome, inefficient, and often fail to account for occlusion, sensitivity to thin objects, and display instability, leading to energy wastage, particularly in battery-operated devices.
Innovation Solution
The system includes a computer system with a display generation component, cameras, and input devices, utilizing methods to determine occlusion and spatial relationships, and provide intelligent visualization and manipulation of virtual objects, reducing user inputs and conserving power.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional methods display all portions of virtual objects, then complete visualization is achieved, but occlusion by physical objects is not properly represented and energy is wasted
Solution Approach 1:
The system extracts and identifies portions of virtual objects that are occluded by physical objects using depth sensing and spatial analysis, then removes these occluded portions from rendering. This selective extraction approach prevents wasting energy on displaying invisible portions while maintaining accurate occlusion representation.
Solution Approach 2:
The system performs preliminary spatial analysis and occlusion detection before rendering virtual objects. By pre-determining which portions will be occluded based on physical object positions and depths, the system avoids unnecessary rendering computations and energy consumption.
2Measurement precision
If conventional methods use simple boundary detection, then processing speed is maintained, but uncertainty in occluding object boundaries is not accounted for
Solution Approach 1:
The system applies different detection strategies to different regions of physical objects based on local characteristics. High-precision depth sensing is applied to regions with uncertain boundaries or significant occlusion impact, while simpler methods are used where boundaries are clear, optimizing both precision and complexity.
Solution Approach 2:
The system introduces depth sensors and spatial mapping as intermediary components between the camera and the rendering engine. These intermediaries provide precise boundary detection data that resolves uncertainty in occluding object boundaries without requiring complex post-processing.
3Stability of the object's composition
If conventional methods render intermediate states, then visual stability is achieved, but processing time increases and energy is wasted
Solution Approach 1:
The system implements periodic stability checks rather than continuous rendering of intermediate states. By periodically assessing whether the virtual object's position and orientation have stabilized, the system maintains visual stability without the continuous processing overhead of rendering every intermediate state.
Solution Approach 2:
The system renders only the necessary portions of intermediate states that contribute to visual stability, rather than fully rendering all intermediate transformations. This partial action approach maintains perceived stability while reducing processing time and energy consumption.
4Ease of operation
If conventional methods continuously process user inputs, then responsiveness is maintained, but energy consumption increases in battery-operated devices
Solution Approach 1:
The system implements event-driven processing where the augmented reality interface only processes user inputs when actually triggered. Rather than continuous polling, the system waits for user actions to occur, then processes them immediately, maintaining responsiveness while eliminating unnecessary continuous processing that wastes battery power.
Data Source
AI summary
A computer system having one or more cameras displays in an augmented reality user interface a representation of a field of view of the one or more cameras, including a plurality of objects in a physical environment. In response to one or more first user inputs, the system places or moves a virtual object at or to a location in the representation of the field of view that corresponds to a physical location on or near a first surface of a first physical object. If the virtual object is positioned at a portion of the first surface that does not include other physical objects, or that includes a physical object that extends from the first surface by less than a threshold amount, the first virtual object is in the representation of the field of view with a predefined spatial relationship to a representation of the first surface.