Mission planner, autonomous flight and precision landing of the Parrot ANAFI drone using Kalman filter

This project, part of my bachelor thesis, creates a simulation model for the precise autonomous landing of a multirotor UAV on a moving platform using Aruco fractal markers and ultra-wideband technology,

A deep learning–based system for Terrain Relative Navigation (TRN) that detects lunar craters using a YOLOv8 neural network and estimates crater depth and inter-crater distances for spacecraft localization.

Full-stack autonomous UAV landing on a moving UGV using PX4 SITL, ROS 2, and Gazebo. Combines multi-scale AprilTag bundle perception, multi-sensor EKF state estimation, and FOV-constrained linear MPC, orchestrated via a reactive Behavior Tree.

In this project we are comparing the 7 different segmentation model to compare their number of parameters accuracy, inference time, training epochs under the same training conditions

Vision-based autonomous quadrotor landing on moving ground rovers and 6-DOF seakeeping ship decks (MuJoCo) - an onboard-sensor-only stack with geometric/MPC/IBVS/RL control, formal-safety shields (HJ reachability + control-barrier functions), maritime fidelity, and a decentralized multi-drone swarm module.