Development of an UAV System Architecture Using Machine Vision Algorithms with a Single-Board Computer or FPGA

This paper presents the development of an unmanned aerial vehicle (UAV) control system architecture based on a single-board computer or FPGA with a neural network. The main focus is on integrating neural networks into UAV flight controllers to enhance control efficiency and reliability. The paper describes the methods and approaches to implementing such a system, as well as analyzes potential benefits and issues that may arise with its use. The outcomes of this research may benefit scientists and engineers in the fields of avionics and artificial intelligence and contribute to the advancement of UAV technologies.