Autonomous Device Self-Position Estimation Using Disturbance-Resistant Landmark Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-position estimation methods for autonomous moving devices are prone to errors and accuracy deterioration due to external disturbances like obstacles and lighting changes, especially when landmarks are arranged on a single plane, such as the ceiling, limiting their robustness in diverse environments.
Innovation Solution
An autonomous moving device equipped with a wide-angle camera, standard cameras capable of posture control, and internal sensors, which includes an odometry unit, base information extraction, landmark selection, tracking, and self-position estimation units to identify and track landmarks unaffected by external disturbances, using a database to extract and manage external disturbance factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If landmarks are arranged on a single plane (ceiling) and detected using wide angle camera, then self position estimation can be performed without needing multiple landmarks, but the system becomes vulnerable to external disturbances such as shielding obstacles and lighting changes
Solution Approach 1:
The patent transitions from two-dimensional ceiling landmarks to three-dimensional vertical landmarks on building surfaces. By adding the vertical dimension and using standard cameras that can control posture to track these landmarks, the system achieves more robust self-position estimation that is less susceptible to external disturbances while maintaining simplified landmark arrangement requirements.
2Area of stationary object
If wide angle camera is used to capture ceiling landmarks, then field of view becomes wide and multiple landmarks are not needed, but accuracy deteriorates under external disturbances such as obstacles and lighting changes
Solution Approach 1:
The patent introduces standard cameras capable of posture control as intermediaries between the wide angle camera's wide field of view and the precision requirements for self-position estimation. The standard cameras actively track selected landmarks, providing stable, high-precision measurements that are less affected by external disturbances while benefiting from the wide field of view provided by the wide angle camera.
3Reliability
If landmark candidate is detected from wide range on ceiling using wide angle camera, then self position can be estimated robust to shielding obstacles, but the system cannot handle various external disturbances such as lighting changes and moving obstacles
Solution Approach 1:
The patent makes the system dynamic by introducing posture-controlled standard cameras that can actively adjust and track landmarks. This dynamic capability allows the system to adapt to various external disturbances including lighting changes and moving obstacles, while maintaining the robustness against shielding obstacles provided by the wide field of view.
Data Source
AI summary
An object of the present invention is to provide an autonomous moving device or the like that estimates a self position based on an arrangement of landmarks in surrounding environment while suppressing the influence by external disturbances. In order to achieve the object described above, an external disturbance factor which becomes an external disturbance and a landmark used for the self position estimation when estimating a self position from an image captured by a wide angle camera are extracted, the landmark which is not affected by the external disturbance factor is selected from an image captured by the standard camera capable of controlling a posture using the extracted information, the posture of the standard camera is controlled so as to track the selected landmark, and the self position is estimated based on the arrangement of the landmarks captured by the tracking-controlled standard camera.