University of Twente Student Theses

Login

UAV-based data streaming for remote processing and real-time information retrieval

Bouma, M.C. (2025) UAV-based data streaming for remote processing and real-time information retrieval.

[img] PDF
8MB
Abstract:Unmanned Aerial Vehicles (UAVs), commonly known as drones, are increasingly used in a variety of applications. This thesis explores if 4G/5G would be a viable option to send data from the drone to a server/workstation where the data can be used for further data processing. In this thesis, the estimation of a vehicle’s speed was considered as use case. A system was designed to send images with Real-Time Streaming Protocol (RTSP) and telemetry data using the Message Queuing Telemetry Transport (MQTT) protocol from the drone to a server/worksta- tion. This RTSP method is reliable and has a latency (time needed to send and fully receive the image) of around 0.45 seconds for a data stream of 10 FPS with an image size of 724x1280 pixels. These received images and telemetry data are then used in algorithms to estimate the speed of the vehicles that can be seen in the images. This is done by making a 3D map with ORB-SLAM,3 which is connected with 3D GPS coordinates, so a sparse point cloud is created, and these 3D points are in GPS coordinates. Then, the vehicle’s position is estimated by detecting the objects with YOLO and tracking these objects with a Bot-SORT tracker. The pixel values of these vehicles are converted to 3D points with the help of the 3D map, and these 3D locations are then used to estimate the speed of the vehicle. This implementation is expandable with multiple drones, and the map creation can even be combined with the help of ORB-SLAM3. Experimental validation at 120 meters altitude demonstrates that the system that was designed is ca- pable of sending images and telemetry data using 4G/5G. This system proposes a scalable and flexible alternative to traditional static infrastructure, enabling real-time traffic analysis and other remote sensing applications.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:38 earth sciences, 54 computer science
Programme:Robotics MSc (60973)
Link to this item:https://purl.utwente.nl/essays/106103
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page