Autonomous mobile robotics is one of the fields that have seen consistent development in recent years. An autonomous mobile robot must be able to move independently of humans in the environment in which it is used; to do so, one of the tools is the availability of a map of the environment. One of the main tasks of these robots is the exploration, which consists in the incremental construction of the map of an environment in which the robot moves. In some situations, for example in the case in which there are closed doors, it might happen that the environment is not completely explorable by the robot. The aim of this thesis is to develop a method that allows a robot to manage these situations, by predicting the possible appearance of the rooms hidden behind closed doors. Starting from a map generated during the exploration of an environment with closed doors that hide rooms, our approach estimates and predicts the structure of the hidden rooms such that they are as similar as possible to their real counterparts. We evaluate the performance of our method on different simulated environments, and also on some real environments.
La robotica mobile autonoma è uno dei campi che hanno visto un consis- tente sviluppo in questi ultimi anni. Un robot mobile autonomo deve essere capace di muoversi nell’ambiente in cui è utilizzato indipendentemente dagli operatori umani; per fare questo uno degli strumenti a sua disposizione è la presenza di una mappa dell’ambiente. Uno dei compiti principale di questi robot è quindi l’esplorazione, che consiste nella costruzione incrementale della mappa di un ambiente in cui il robot si muove. Può succedere, però, che, in alcune situazioni, come per esempio quando sono presenti porte chiuse, l’ambiente non sia completamente esplorabile dal robot. L’obiettivo di questa tesi è lo sviluppo di un metodo che permetta ad un robot di gestire queste situazioni, predicendo il possibile aspetto geometrico delle stanze nascoste da porte chiuse. Partendo da una mappa generata durante l’esplorazione in un ambiente sconosciuto con diverse porte chiuse che nascondono stanze, il nostro approccio predice la struttura delle stanze nascoste in modo tale che siano il più possibile simili alle loro controparti reali. Valutiamo le prestazioni del nostro metodo, su numerosi ambienti simulati e su un limitato numero di ambienti reali.
A method for predicting the layouts of hidden rooms in robotic mapping
AMADELLI, FEDERICO
2018/2019
Abstract
Autonomous mobile robotics is one of the fields that have seen consistent development in recent years. An autonomous mobile robot must be able to move independently of humans in the environment in which it is used; to do so, one of the tools is the availability of a map of the environment. One of the main tasks of these robots is the exploration, which consists in the incremental construction of the map of an environment in which the robot moves. In some situations, for example in the case in which there are closed doors, it might happen that the environment is not completely explorable by the robot. The aim of this thesis is to develop a method that allows a robot to manage these situations, by predicting the possible appearance of the rooms hidden behind closed doors. Starting from a map generated during the exploration of an environment with closed doors that hide rooms, our approach estimates and predicts the structure of the hidden rooms such that they are as similar as possible to their real counterparts. We evaluate the performance of our method on different simulated environments, and also on some real environments.File | Dimensione | Formato | |
---|---|---|---|
tesi.pdf
accessibile in internet per tutti
Descrizione: Thesis text
Dimensione
11.04 MB
Formato
Adobe PDF
|
11.04 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/165219