A realistic simulation platform for multi-quadcopter search using downward facing cameras

Abstract

This paper presents a cross platform simulation environment, in a hybrid centralized-decentralized architecture, for multi-quadcopter search problem developed using MATLAB and Gazebo simulator in Robot Operating System (ROS) environment. Multiple quadcopters equipped with downward facing camera are deployed in a search area to gather information such as presence of targets of interest. Search is modeled as reducing uncertainty density, a metric of lack of information about the presence (or absence) of the targets of interest. The simulation platform developed will be a very useful tool for conducting realistic simulation experiments to validate the proposed search strategy and to make a comparative study of its performance in terms of time for the search process, with parameters such as camera search effectiveness models, sensor range, the number of robots, to decide on the parameters best suited for a given situation. We provide simulation results demonstrating the proposed search strategy using the developed simulation environment. © 2019