Visual-Inertial Device Tracking via Marker and Scene Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented reality technologies face challenges in accurately tracking devices, such as head-mounted displays and smartphones, within indoor environments due to the lack of reliable methods for determining their position and pose relative to markers, especially when markers are not consistently visible.
Innovation Solution
A method utilizing a terminal device equipped with a first camera, a second camera, and an inertial measurement unit (IMU) to acquire information from images of markers and target scenes, combining visual and inertial data to determine the device's position and pose relative to markers, even when markers are not visible, through visual-inertial odometry (VIO) and marker-based tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If marker-based tracking is used, then tracking accuracy is improved when markers are visible, but tracking reliability deteriorates when markers are not consistently visible
Solution Approach 1:
The patent combines marker-based tracking with visual-inertial odometry (VIO) into a unified tracking system. When markers are visible, the system uses marker-based tracking for high precision; when markers are not visible, it seamlessly transitions to VIO using the second camera and IMU to maintain continuous tracking reliability. This merging of two different tracking methods resolves the contradiction between accuracy and reliability.
Solution Approach 2:
The patent introduces an intermediary system (VIO using second camera and IMU) that acts as a backup and supplement to marker-based tracking. This intermediary tracking method ensures that tracking reliability is maintained even when the primary marker-based method fails, thus resolving the reliability issue without compromising the accuracy benefits of marker tracking when available.
2Duration of action of stationary object
If multiple sensors (first camera, second camera, IMU) are used for visual-inertial odometry, then tracking continuity is improved, but device complexity increases
Solution Approach 1:
The patent segments the tracking system into two distinct functional parts: a first camera dedicated to marker detection and a second camera paired with IMU for VIO. This segmentation allows each component to have specialized functions, improving tracking continuity while managing complexity through clear division of labor rather than requiring all sensors to perform all functions.
Solution Approach 2:
The patent implements a multi-functional tracking system where the first camera can work independently for marker tracking, the second camera can work independently for VIO, and they can also work together. This universality allows the system to maintain tracking continuity across different scenarios without requiring complex integration, as each component can operate autonomously or in combination based on conditions.
Data Source
AI summary
A method of device tracking is provided. Based on a captured image containing a marker, first spatial position is acquired. Based on a captured image of a scene, second spatial position is acquired. Based on at least one of the first spatial position and the second spatial position, a terminal device may be positioned and tracked.