A SELF-ORGANIZED AUTOMATED SYSTEM TO CONTROL UNMANNED AERIAL VEHICLES FOR OBJECT DETECTION

Abstract

Acquiring images dynamically in a three-dimensional space and subsequently processing them Acquiring images dynamically in a three-dimensional space and subsequently processing them to identify precise structural objects is crucial. Ensuring high recognition accuracy and proper, comprehensive image definition is vital. It is also imperative to incorporate detection features within the self-organized system during input data classification. In the present study, we propose a unique self-organized automated system where one or multiple UAVs are managed and tracked to capture images of identified objects, examining one object at a time. The developed architecture of an automated system for the dynamic acquisition of images of structural objects in three-dimensional space allows for reaching the appropriate level of organization when determining the next steps in the functioning of subsystems and components. Control tools provide programmatic mission control by grouping a fixed number of UAVs into groups and performing targeted work in fragments of the operating environment. The monitoring software module of the automated system processes the mission output data. It analyzes and compares based on already valid data to ensure the most accurate result of calculations of the number of fruits on the trees. The findings of this research lay the foundation for developing new tools capable of launching and supervising unmanned aerial vehicles over subsets of the analyzed spatial region based on the provided initial data. The devised architecture facilitates reaching the suitable level of organization when deciding the following actions for operating subsystems and components. The performed experiments verify the feasibility of implementing the suggested architectural approaches. Directions for further research include improving the flight methods implemented in the system, image recognition, and calculating the number of recognized objects.