Mapiranje prostora korištenjem senzora Pepper robota

undergraduate thesis

• Title (english): Mapping of environment with Pepper robot sensors
• Author: Karlo Valentin Cihlar
• Mentor: Zdenko Kovačić (mentor)
• Committee member: Zdenko Kovačić (predsjednik povjerenstva)
• Committee member: Juraj Oršulić (član povjerenstva)
• Committee member: Ivan Hrabar (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-07-09, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Kako bi robot Pepper mogao obavljati funkciju welcome managera na Fakultetu elektrotehnike i računarstva, potrebno je izgraditi mapu zauzeća prostora fakulteta kako bi se robot mogao efikasno lokalizirati i navigirati u prostoru. U tu svrhu uspoređena su 3 različita postupka mapiranja korištenjem Pepperovih senzora: mapiranje korištenjem lasera, mapiranje korištenjem dubinske kamere i mapiranje korištenjem fuzije oba senzora. Rezultati pokazuju da je mapiranje laserima za bilo kakav složeniji i veći prostor veoma loše kvalitete te je nedostatno za primjenu u željenoj funkciji robota. Uzrok tome je mala gustoća lasera i pogreška u rotacijskoj odometriji. Mape dobivene korištenjem 3D kamere zadovoljavajuće su kvalitete i mogle bi se koristiti u funkciji Peppera kao welcome managera. Kod mapa dobivenih fuzijom pojavljuje se dodatno rasipanje zbog istovremenog korištenja lasera i dubinske kamere. Mape su slične onima dobivenim iz dubinske kamere, uz dodatnu pogrešku. Iako bi i ovaj način mapiranja u nekim primjenama bio dostatne kvalitete, predlaže se da se u funkciji Peppera kao welcome managera koriste mape dobivene dubinskom kamerom, zbog jednostavnijeg postupka mapiranja i veće točnosti mape.

Abstract (english)

In order for the Pepper robot to be able to perform the function of a welcome manager at the Faculty of Electrical Engineering and Computing, it is necessary to build a map of the faculty interior so that the robot can be localized and navigated in space. For this purpose, 3 different mapping procedures using Pepper sensors were compared: mapping using lasers, mapping using a depth camera and mapping using the fusion of both sensors. The results show that mapping with lasers for a bit more complex, larger spaces is of insufficient quality for application in the desired robot function. This is caused by the low density of the laser scans and the error in the rotational odometry. The maps obtained using the 3D camera are of satisfactory quality and could be used with Pepper functioning as a welcome manager. In fusion maps, additional scattering occurs due to the simultaneous use of lasers and a depth camera. The maps are similar to those obtained from a depth camera, with added error. Although this way of mapping would be of sufficient quality in some applications, it is suggested that in the Pepper's function as a welcome manager, maps obtained with a depth camera be used, due to the simpler mapping process and greater accuracy of the map.

• Keywords: robot Pepper; ROS; mapiranje; dubinska kamera; laseri
• Keywords (english): Pepper robot; ROS; mapping; depth camera; lasers
• Language: croatian
• URN:NBN: urn:nbn:hr:168:206918
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka elektrotehnike i informacijske tehnologije (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 14:40:46