| Abstrakt: | This bachelor's thesis describes the development of the iOS application CR Fly, which is designed to enhance the process of 3D scene reconstruction using images captured by DJI drones. The application utilizes the DJI Software Development Kit (SDK), which enables bidirectional communication with the drone and RealityCapture software, allowing the reconstruction of real-world objects into their digital twins using images obtained from drones. The development of the application was carried out with an emphasis on modularity and scalability using the Model-View-Controller architecture in the Swift programming language. The application also provides the user with a real-time view from the drone's camera, as well as analysis and 3D reconstruction of the captured environment, significantly aiding the pilot in capturing individual objects. Extensive testing confirmed the functionality of the application and its ability to reconstruct detailed 3D models of various scenes. For experimental reconstruction, the Monument of Nicolaus Copernicus and the Church of St. Elizabeth of Hungary were selected, representing simple and complex scenes, which serve to verify the practical functionality of the application. To address further limitations of existing solutions, specifically autonomous flights, the application was designed to be extendable in the future with a module for autonomous scanning. This extensibility is enabled by the mobile device already possesing photographs, camera positions, 3D reconstructions of the scene, and additionally, the DJI SDK allows setting waypoints for the drone's autonomous flight.
|
|---|