Development of an autonomous target interception package that allows a drone to move towards a target using vision based navigation
With the increasing threat of hostile drone swarms, there needs to be a solution to neutralise these threats. One of such solutions is the use of a group of drone interceptors. These interceptors require a target interception software package that will guide them towards their assigned targets before they can perform neutralisation action.
Sensor: Azure Kinect DK Stereo Camera
Object Detection and Classification: YOLOv8m
Guidance Law: Image-Based Visual Servoing
Velocity Controller: PID
Flight Controller: ArduPilot
Simulator: Gazebo
Simulated drone: Iris SITL
Implementing gimbal control to enable more robust target locking.
roslaunch ardupilot_gazebo multi_iris_with_roscam.launch
sim_vehicle.py -v ArduCopter -f gazebo-drone1 --console -I0 --out=tcpin:0.0.0.0:8100
sim_vehicle.py -v ArduCopter -f gazebo-drone2 --console -I1 --out=tcpin:0.0.0.0:8200 -M
roslaunch iq_sim multi_apm.launch
roslaunch iq_sim multi_cuav_gnc.launch
roslaunch iq_sim cuav_interceptor_gnc.launch
roslaunch iq_sim cuav_interceptor_gnc_gimbal.launch