AR Target Positioning via Multi-Camera Vector Intersection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current augmented reality systems lack efficient methods for scenario triggering and interaction based on target positioning, which is essential for seamlessly integrating computer-generated content with real-world environments.
Innovation Solution
A system and method that utilize multiple camera devices to capture images, determine target object positions by computing intersection vectors, and compare these positions with predefined task triggering regions to initiate specific tasks or content presentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple camera devices are used to capture images and compute intersection vectors for target positioning, then measurement precision of target object position is improved, but device complexity increases
Solution Approach 1:
The system divides the positioning task into multiple independent camera devices, each capturing images from different positions. Each camera independently identifies the target object and computes its own target object vector, with the final position determined by intersecting these vectors. This segmentation allows parallel processing and improves measurement precision without requiring a single complex device.
Solution Approach 2:
The patent introduces a triggering controller as an intermediary that receives images from multiple camera devices, performs target object identification, computes intersection vectors, and determines the final position. This intermediary coordinates the multiple cameras and manages the complex computations, isolating the complexity from the camera devices themselves while maintaining high measurement precision.
2Measurement precision
If target object position is determined by computing intersection of vectors from multiple camera positions, then measurement precision is improved, but computing time increases
Solution Approach 1:
The system pre-computes and stores extrinsic parameters (rotation and translation matrices) between multiple camera positions before actual target positioning is needed. This preliminary preparation of coordinate transformation data allows the intersection computation to proceed quickly using pre-established geometric relationships, reducing real-time computing time while maintaining precision.
Solution Approach 2:
The patent implements a triggering mechanism that activates the full multi-camera positioning system only when a target object enters a predefined triggering region. During normal operation, the system performs lighter monitoring tasks. This partial action approach reduces average computing time by avoiding continuous full-power computation while maintaining high precision when actually needed.
Data Source
AI summary
A method for scenario triggering and interaction is disclosed. One or more camera devices (110) are used to monitor a target object (142) in a scene (140). A target object vector (143a) is generated from captured images of each camera device (110) to represent a line of sight from the camera device (110) to the target object (142). A target object position (129) in the scene (140) is then determined from an intersection of the target object vectors (143a). The target object position (129) is compared with a task triggering region (144) in the scene (140) to trigger a pre-determined task when the target object (142) is within or around the task triggering region (144). For example, the target object (142) may be a human user and augmented reality contents are presented to the human user when the human user is detected to be within or around the task triggering region (144).