Ceiling Camera Vehicle Localization via Robot Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle localization systems face challenges in determining the location of vehicles within structures like warehouses or parking garages, where GPS signals are unavailable and sensor data may be obstructed or affected by lighting conditions.
Innovation Solution
A vehicle locator system that includes network nodes with downward-facing cameras mounted to a ceiling, which capture images of vehicles on the floor. A mobile robot with an upward-facing camera is used to calibrate the system by determining the poses of the cameras through visual odometry and fiducial marker detection, allowing the system to accurately predict vehicle locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS satellite signals are used for vehicle localization, then localization accuracy is improved, but the system becomes unavailable in structures such as warehouses and parking garages where satellite signals are blocked
Solution Approach 1:
The patent introduces an intermediary calibration system consisting of stationary cameras mounted on the ceiling and a mobile robot with an upward-facing camera. This intermediary system captures images to determine poses of stationary cameras, creating a bridge that enables vehicle localization through image processing rather than direct GPS satellite signals, thereby resolving the contradiction between localization accuracy and system availability in GPS-denied environments
Solution Approach 2:
The patent replaces the GPS satellite signal-based mechanical/electromagnetic system with an image-based visual localization system. By using downward-facing stationary cameras and upward-facing mobile robot cameras to capture images and determine poses through image processing, the system substitutes the GPS dependency with a visual measurement system that operates independently of satellite signals
2Reliability
If sensor data from cameras is used for vehicle localization, then the system can operate without GPS, but data becomes unavailable or unreliable due to lighting conditions and obstructions
Solution Approach 1:
The patent transitions from traditional horizontal/ground-level camera views to a vertical dimension by mounting stationary cameras on the ceiling facing downward and using a mobile robot with an upward-facing camera. This dimensional change allows the system to capture images from overhead perspectives, reducing the impact of ground-level obstructions and lighting conditions on measurement precision while maintaining system availability
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
The system effectively determines the location of vehicles within structures, enabling accurate navigation and actuation of vehicles, even in environments where traditional localization methods fail.
Implementation Method 1
determining upward-facing camera poses at respective times based on the upward images; and determining respective poses of the downward-facing cameras based on (a) describing motion of the robot from the downward images, and (b) the upward-facing camera poses determined from the upward images
Implementation Method 2
capturing downward images with the respective downward-facing cameras; determining upward-facing camera poses at respective times based on the upward images; and determining respective poses of the downward-facing cameras based on (a) describing motion of the robot from the downward images
Data Source
AI summary
A robot can be moved in a structure that includes a plurality of downward-facing cameras, and, as the robot moves, upward images can be captured with an upward-facing camera mounted to the robot. Downward images can be captured with the respective downward-facing cameras. Upward-facing camera poses can be determined at respective times based on the upward images. Further, respective poses of the downward-facing cameras can be determined based on (a) describing motion of the robot from the downward images, and (b) the upward-facing camera poses determined from the upward images.