Drone Camera Orientation for GPS-Free Visual Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Drones using autonomous flight modes face challenges in maintaining accurate self-position estimation due to the inability to detect feature points in certain environments, leading to prolonged travel times and reduced efficiency.
Innovation Solution
A mobile body control device and method that controls the imaging direction of a camera to focus on areas where localization is feasible, using localization feasibility information to guide the drone along paths where self-position estimation can be accurately performed, even without external position information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the drone performs autonomous flight using SLAM process, then self-position estimation can be performed without external position information, but self-position estimation cannot be performed in areas where no feature points can be detected (e.g., white walls)
Solution Approach 1:
The system performs preliminary mapping and pre-identifies localization feasible areas before autonomous flight. By storing this feasibility information in advance, the drone can plan its flight path to stay within areas where feature points are detectable, preventing localization failures before they occur.
Solution Approach 2:
The system continuously monitors localization feasibility during flight and dynamically adjusts the flight path based on real-time feedback. When approaching areas with low localization feasibility, the system redirects the drone to maintain high estimation reliability, creating a closed-loop control system.
2Measurement precision
If the drone moves to areas where self-position estimation with high reliability can be performed, then localization accuracy is improved, but the time required to reach destination becomes significantly long due to repeated movements
Solution Approach 1:
The system pre-identifies and stores localization feasible areas before flight begins. This allows the flight path planning to occur offline, finding optimal routes that maintain high localization accuracy without requiring real-time path adjustments, thus minimizing travel time.
Solution Approach 2:
The system introduces localization feasibility information as an intermediary layer between the start and destination points. This intermediary data structure enables the planning unit to compute paths that naturally pass through high-reliability areas without requiring repeated detours during actual flight.
3Productivity
If the drone follows a direct route to destination, then travel time is minimized, but self-position estimation accuracy deteriorates when passing through areas with many low-reliability localization zones
Solution Approach 1:
The localization feasibility information acts as an intermediary that mediates between the direct route requirement and localization accuracy. The planning unit uses this intermediary data to compute a path that is nearly direct but strategically routed through high-feasibility areas, achieving both efficiency and accuracy.
Solution Approach 2:
The system changes the path planning parameter from pure shortest distance to a composite metric that incorporates localization feasibility. By weighting both distance and localization quality, the system generates paths that maintain high efficiency while ensuring sufficient position estimation accuracy throughout the flight.
Data Source
AI summary
A device and a method realizing movement according to a pre-defined path even in a case that a mobile body cannot input absolute position information from the outside such as a GPS signal or the like are provided. There is provided a configuration executing control of a mobile body such as a drone or the like, and the configuration executes an imaging direction control step of controlling an imaging direction of a camera using a control unit and a localization processing step of executing a localization process estimating a self-position using a captured image of the camera using a self-position estimating unit. In the imaging direction control step, a camera imaging direction control process of directing the imaging direction of the camera to a divisional area that can be localized by referring to localization feasibility information enabling to identify whether or not localization can be performed in units of divisional areas is executed.