AGV sledující pohyb osoby

Abstract

This thesis deals with the complex task of design and implementation of the automated guided vehicle (AGV). The guiding method this vehicle uses is near-distance person detection and localization. Ultra-wideband localization technology (UWB) is used as the main detection method, supplemented with image-processing and object-detection using an RGBD camera. In the case of image processing, multiple detection methods are used and compared. Part of the work is the creation of control software, as well as the design and construction of AGV hardware. Within this thesis, two control algorithms were proposed and possibilities for further development were outlined. In addition, a simulation was created to allow the robot to be tested before the construction of the robot itself was completed.