AR Headset Drift Correction via Downward Camera
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Augmented Reality (AR) devices, particularly Head-Mounted Displays (HMDs) with downward-facing displays, experience sensor drift and incorrect placement of virtual objects due to saturated sensors, leading to inaccuracies in orienting virtual objects relative to the real environment.
Innovation Solution
Incorporating a rear-mounted camera and additional optics to provide a reference perspective on the environment, allowing the AR device to correct sensor data and accurately position virtual objects by mapping real objects and using a method that captures a station keeping image to determine the user's field of view and anchor virtual objects accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If downward-facing display and sensor are used in HMD, then virtual objects can be projected onto real environment, but sensor drift occurs and virtual object placement becomes inaccurate
Solution Approach 1:
The system captures images of the real environment through the downward-facing camera, detects real objects and their positions, and uses this visual feedback to correct sensor drift. The detected real object positions serve as reference points to continuously calibrate and adjust the virtual object placement, ensuring long-term accuracy despite sensor instability.
Solution Approach 2:
Real objects in the environment act as intermediary reference points between the sensor and the virtual objects. The system uses these real objects as stable anchors to mediate the relationship between sensor data and virtual object positioning, providing a reliable reference frame that compensates for sensor drift.
2Measurement precision
If additional camera and optics are added to correct sensor drift, then virtual object placement accuracy improves, but device complexity increases
Solution Approach 1:
The downward-facing camera serves multiple functions: it captures the real environment for drift correction, provides visual feedback for object detection, and enables augmented reality display. This multi-functionality reduces the need for separate dedicated components, managing device complexity while maintaining measurement precision.
Solution Approach 2:
The system uses its own downward-facing camera and processor to perform drift correction and environmental mapping, making the device self-sufficient. The mobile computing device's existing hardware resources are leveraged to provide correction capabilities without requiring external correction systems.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This solution effectively reduces sensor drift and improves the accuracy of virtual object placement within the real environment, enhancing the AR experience by leveraging the hardware resources of mobile computing devices and reducing the need for dedicated computer processors.
Implementation Method 1
an optical arrangement configured to receive light generated by the screen and reflect the light onto the viewport for reflection to a wearer of the AR device
Implementation Method 2
an additional optic, arranged above the camera to direct light from the environment outside of a field of view of the wearer to the camera
Data Source
AI summary
Embodiments provide for tracking location and resolving drift in Augmented Reality (AR) devices. The AR devices includes computing devices having screens on a first face and cameras on a second, opposite face to project an image onto optical arrangements for viewing by wearers of the AR devices. The AR devices map locations for real objects in the environment to a virtual environment; anchor virtual objects at anchor locations within the virtual environment; capture station keeping images of the environment from a first Field of View via the camera; determine a second, different Field of View in the environment for the wearer of the AR device based on the relative locations of real objects present in the station keeping images; and output images depicting the virtual objects at positions on the screen to depict the virtual objects in the physical environment at the anchor locations.