Virtual Object Positioning in Moving Vehicles Using Segmented Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented reality (AR) technologies face challenges in maintaining accurate positioning of virtual objects within both moving interior and exterior environments of a vehicle, as inside-out tracking techniques are disrupted by differing speeds of real-world objects inside and outside the vehicle.
Innovation Solution
A method and system that capture images of both interior and exterior environments, classify them into segments, and use separate localization models to estimate poses for virtual object placement, allowing accurate display of virtual objects within the user's field of view despite vehicle movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If inside-out tracking technique is used for self-positioning in moving vehicles, then the HMD can perform positioning without fixed lighthouses, but real-world objects inside and outside the vehicle moving at different speeds disturb the positioning process, causing the HMD to fail to display virtual objects at correct positions
Solution Approach 1:
The patent segments the image into interior space and exterior environment portions, then applies separate localization models to each segment. This allows independent tracking of objects inside the vehicle (which move with the vehicle) and objects outside the vehicle (which move relative to the vehicle at different speeds), resolving the positioning disturbance caused by combining both environments in a single tracking model.
2Measurement precision
If outside-in tracking technique with fixed lighthouses is used, then high accuracy and simple algorithms are achieved, but the fixed lighthouses are not applicable to vehicles that are often in a moving status
Solution Approach 1:
The patent implements inside-out tracking where the HMD performs self-positioning by capturing and processing images of the surrounding environment itself, without requiring external fixed lighthouses. The device uses its own camera to capture images, processes them through localization models, and determines its position and orientation autonomously, making it suitable for moving vehicles where external reference structures are unavailable.
Data Source
AI summary
A method for providing a virtual environment during movement is provided. The method includes the following operations: capturing a first image associated with an interior space of a housing and also associated with part of an external environment captured outward from the interior space; classifying the first image into a first segment associated with the interior space and a second segment associated with the part of the external environment; estimating a first pose and a second pose of a mobile device associated with respective the housing and the external environment, in which the first pose is estimated by a first localization model based on the first segment, and the second pose is estimated by a second localization model based on a second image associated with the external environment; and displaying virtual objects by the mobile device according to the first and second poses.