Visual Target Detection Influenced Trajectory Generation

Implemented using path planning algorithm from A path planning algorithm for a crop monitoring fixed-wing unmanned aerial system that develops a path in a field to find an object, which is detected using a drone camera feed that is processed with OpenCV and YOLOV10. Once the object is detected a trajectory towards it is generated. The recommended path, along with current drone statistics and live camera feed are displayed in a HUD.

Run Instructions

• Create an Anaconda environment
  • conda create -n target_tracking
  • conda activate target_tracking
• Run pip install -r requirements.txt
• Start the "ground control" with streamlit run main.py

Individual Script Testing

• Run python3 path_planning.py to generate Waypoints given the input vector map
• Run python3 object_detection.py to detect objects given a camera feed - requires Waypoints as input
• Run python3 trajectory_generation.py to generate heading commands and a waypoint