This thesis addresses the development of a control and communication platform for Crazyflie mini-drones within the UAV navigation framework Aerostack 2, currently being developed by the Computer Vision and Aerial Robotics (CVAR) group of the Center of Automation and Robotics (CAR), for autonomous missions in the field of indoor logistics. For this, it is necessary that the developed platform is integrated with all the existing functionalities within Aerostack 2, as well as the development of new modules necessary for its correct operation. In addition, a use case for the identification of europallets using computer vision and deep learning techniques is proposed. The Crazyflie AI Deck expansion module incorpo- rates an RGB camera and Wifi connectivity that can be used for live image transmission. These images are subsequently processed by a pallet identification system based on work contributed by the Technical University of Dortmund. A YOLO model is used for the detection of europallet support blocks and subsequent pallet identification using a Part- based Convolutional Baseline (PCB) architecture. As the thesis is the result of a collaboration between the Polytechnic University of Madrid and the Technical University of Dortmund, most of the development and testing has been carried out at the TUD facilities in Germany .
This thesis addresses the development of a control and communication platform for Crazyflie mini-drones within the UAV navigation framework Aerostack 2, currently being developed by the Computer Vision and Aerial Robotics (CVAR) group of the Center of Automation and Robotics (CAR), for autonomous missions in the field of indoor logistics. For this, it is necessary that the developed platform is integrated with all the existing functionalities within Aerostack 2, as well as the development of new modules necessary for its correct operation. In addition, a use case for the identification of europallets using computer vision and deep learning techniques is proposed. The Crazyflie AI Deck expansion module incorpo- rates an RGB camera and Wifi connectivity that can be used for live image transmission. These images are subsequently processed by a pallet identification system based on work contributed by the Technical University of Dortmund. A YOLO model is used for the detection of europallet support blocks and subsequent pallet identification using a Part- based Convolutional Baseline (PCB) architecture. As the thesis is the result of a collaboration between the Polytechnic University of Madrid and the Technical University of Dortmund, most of the development and testing has been carried out at the TUD facilities in Germany . Read More
Related posts:
AppQuitectura: aplicación móvil para la gamificación en el área de Composición Arquitectónica
Computer Vision-Based Wood Identification: A Review
How visual literacy is developed through Engineering Graphics subjects
A behaviour-based control architecture for heterogeneous modular, multi-configurable, chained micro-robots
Empowering computer vision in higher education: a novel tool for enhancing video coding comprehension
Robot assembly system for computer-integrated construction