Detekcija i praćenje osoba korištenjem Nao robota

undergraduate thesis

• Title (english): Person Detection and Tracking with NAO Robot
• Author: Ante Zglav
• Mentor: Zdenko Kovačić (mentor)
• Committee member: Zdenko Kovačić (predsjednik povjerenstva)
• Committee member: Damjan Miklić (član povjerenstva)
• Committee member: Frano Petric (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2014-07-14, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

U okviru dijagnostike autizma, NAO robot treba detektirati lice djeteta kako bi mogao odrediti pridaje li dijete pažnju robotu. Cilj je zadatka omogućiti NAO robotu da prati osobu u vidnom polju jedne od kamera. Pritom u slučaju da se lice ne detektira treba imati informaciju o tome nalazi li se dijete u vidnom polju kamere ili ne. Kako bi se postigla detekcija ljudi bez korištenja detekcije lica pomoću OpenCV biblioteke implementirane su dvije dodatne metode za detekciju ljudi. Te metode su detekcija pomoću histograma usmjerenih gradijenata te metoda detekcije boje kože. Bilo kojom od te dvije metode može se postići detekcija ljudi ukoliko lice nije pronađeno, ali najbolji rezultati se postižu njihovim kombiniranjem.

Abstract (english)

In the service of diagnostics of autism, a NAO robot will have to detect
the face of the child in order to determine if the child is paying attention to
the robot. The goal of this task is for the robot to track a person within his
field of vision. In case a face is not detected in the field of vision the robot
should still be able to detect wether a person is present or not. In order to
achieve people detection without the use of face detection two methods were
implemented using the OpenCV library. Those methods are people detection
using histograms of oriented gradient and human skin tone detection. Any of
these two methods can detect human beings but in order to achieve the best
results they should be combined.

• Keywords: NAO; detekcija ljudi; HOG; detekcija kože; povratna projekcija; računalni vid
• Keywords (english): NAO; Humandetection; HOG; Skin detection; Back Projection; Computer vision
• Language: croatian
• URN:NBN: urn:nbn:hr:168:134277
• 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: 2018-02-06 14:32:46