Navigation and Self-Semantic Location of Drones in Indoor Environments by Combining the Visual Bug Algorithm and...

Area of Science:

• Robotics
• Computer Vision
• Artificial Intelligence

Background:

• Autonomous navigation in indoor environments presents challenges for robots.
• Visual topological maps offer a framework for robot guidance using landmarks.
• Image entropy can quantify visual information, useful for object detection.

Purpose of the Study:

• To develop a hybrid algorithm for robot self-localization and autonomous navigation.
• To utilize entropy-based vision for landmark detection and obstacle avoidance.
• To validate the proposed algorithm in indoor navigation tasks using an Unmanned Aerial Vehicle (UAV).

Main Methods:

• A hybrid algorithm combining entropy-based vision and visual topological maps is introduced.
• Image entropy is used to identify unique objects as landmarks or obstacles.
• A visual topological map guides robot homing based on detected landmarks.
• The visual bug algorithm is employed for validation with a UAV.

Main Results:

• The algorithm successfully uses image entropy for both landmark detection and obstacle avoidance.
• Visual topological maps guide the robot towards target points (homing).
• The experimental framework validated the algorithm's effectiveness in indoor navigation tasks.

Conclusions:

• The proposed hybrid algorithm enhances robot self-localization and autonomous navigation capabilities.
• Entropy-based vision is a viable method for landmark identification and collision avoidance.
• The approach shows promise for practical applications in indoor robot navigation.