Frontier-Based autonomous 3D Exploration with unmanned aerial vehicles

Implementirani algoritam 3D detekcije fronte, granice istraženog i neistraženog prostora, testiran je u Gazebo i RViz simulatoru na Kopterworx modelu bespilotne letjelice. Pomoću OctoMap biblioteke vizualizirana je izgrađena karta prostora te se pomoću Graph SLAM metode istovremeno ostvaruje dobivanje karte prostora i lokalizacija letjelice. Provedena je analiza dobivenih rezultata na temelju koje je provedena optimizacija metode odabira ciljne točke. Cilj optimizacije je smanjivanje vremena simulacije, tj. istraživanja prostora, uz uvjet da se ne izgube informacije o prostoru. Uspoređen je optimiziran algoritam s postojećim algoritmom. Eksperiment je proveden na virtualnoj simulaciji te je razmotrena primjena algoritma na stvarnom sustavu.The implemented algorithm for 3D frontier detection, the boundary between the explored and the unexplored space, was tested in the Gazebo simulator. The Octomap library is used to visualize the constructed map of the environment and the Graph SLAM method is used to locate a robot and generate a map of the environment. An analysis of the obtained results was performed, on which the optimization of the selection method of the target point was performed. The goal of the optimization is to reduce the simulation time, the time of environment exploration, under the condition that the information about the environment is not lost. The optimized algorithm is compared with the existing algorithm. The experiment was performed in the Gazebo simulator and the application of the algorithm in a real-world setup was considered