Izbjegavanje prepreka korištenjem algoritma potencijalnih polja dobivenih iz slike stereo kamere

undergraduate thesis

• Title (english): Potential field based obstacle avoidance by using data from stereo camera
• Author: Domagoj Šimag
• Mentor: Stjepan Bogdan (mentor)
• Committee member: Stjepan Bogdan (predsjednik povjerenstva)
• Committee member: Frano Petric (član povjerenstva)
• Committee member: Ana Batinović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2019-09-17, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

U radu je obrad¯eno izbjegavanje prepreka u realnom vremenu korištenjem algoritma
umjetnih potencijalnih polja. Kako bi letjelica mogla nesmetano prolaziti prostorom,
potreban joj je senzor. Kao senzor je korištena stereo (RealSense) kamera.
Zahvaljuju´ci karakteristikama realsense kamere, mogu´ce je dobiti dubinsku percepciju
prostora i stvoriti PointCloud. Transformacijom PointCloud-a u OctoMap dobiva
se 3D prikaz prostora. Uzimajuc´i podatke iz izgrad¯ene 3D karte prostora, letjelica
uspješno izbjegava svaku prepreku koja se nalazi na putu prema ciljnoj toˇcki. Osim
navedenog, u radu je objašnjena i mogu´ca primjena genetskih algoritama i umjetnih
neuronskih mreža na gibanje letjelice u prostoru. Eksperimenti su provedeni u simulacijskom
okruženju, a algoritam izbjegavanja prepreka mogu´ce je primijeniti i na
stvarno okruženje.

Abstract (english)

The subject of this paper is real-time obstacle avoidance using artificial potential
fields algorithm. In order to successfully travel through a given area, system needs a
sensor. This paper uses stereo (RealSense) camera. One of the main features of this
camera is depth perception and this feature is used to create PointCloud. By transforming
data from PointCloud to OctoMap, a 3D image of an area is obtained. Using the
data given from 3D image, drone manages to avoid every obstacle in its path. Furthermore,
this paper discusses a possibility of using genetic algorithms and artificial neural
networks in path planning. Experiments were made in simulations and it’s possible to
use this algorithm on real drones.

• Keywords: potencijalna polja; algoritam; octomap; oblak tocaka; stereo kamera
• Keywords (english): potential fields; algorithm; octomap; point cloud; stereo camera
• Language: croatian
• URN:NBN: urn:nbn:hr:168:089400
• 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: 2020-01-07 22:21:14