20.500.12556/DKUM-46481 RAČUNALNIŠKI VID KOT PRIPOMOČEK ZA POUČEVANJE OTROK Z GIBALNIMI MOTNJAMI COMPUTER VISION AS A TOOL FOR TEACHING CHILDREN WITH MOTOR DISORDERS V okviru diplomske naloge smo pripravili svojo rešitev za lažjo uporabo računalnika, za osebe s posebnimi potrebami. Pri tem smo se omejili na mlajše otroke z gibalnimi in lažjimi duševnimi motnjami. Ker želimo, da je naša rešitev čim preprostejša in cenovno dostopna smo za zajem podatkov uporabili cenovno ugodno spletno kamero, celotno aplikacijo pa smo bazirali na konceptu »naprej (naslednja izbira) – potrdi«. Ta koncept nam omogoča, da celotno aplikacijo krmilimo z zaznavanjem dveh vhodnih signalov. Ker je bil cilj naloge ugotoviti katere kretnje so najbolj primerne za krmiljenje računalnika, smo pripravili aplikacijo, ki jo je možno krmiliti z mežikanjem oči, nagibanjem glave, dvigovanjem rok, odpiranjem ust in iztegovanjem jezika. Program smo nato testirali najprej na statičnih slikah in nato še z uporabo spletne kamere. Ko smo bili z delovanjem algoritmov zadovoljni, smo celotno aplikacijo preizkusili še pri ciljnih uporabnikih, torej otrocih z gibalnimi motnjami. Z analizo rezultatov smo nato načine krmiljenja razvrstili po učinkovitosti. Najučinkovitejše se je izkazalo krmiljenje s pomočjo nagibanja glave. In the frame of the thesis, we prepared a solution to facilitate the use of a computer for people with special needs. We limited ourselves to younger children with physical and mild mental disorders. Because we wanted our solution to be as simple and affordable as possible, we used a favourable web camera for capturing data. The entire application however, is based on the concept "forward (next choice) - confirm". This concept allows us to control the entire application by detecting two input signals. Since the objective was to determine which gestures were most suitable for controlling a computer, we created an application that can be controlled by blinking eyes, tilting the head, raising hands, opening the mouth and stretching the tongue. The programme was then tested first on static images and later also by using a webcam. When we were satisfied with the working of the algorithms, we tested the entire application also with the target users, i.e. children with physical disabilities. With the analysis of the results, we ranked the control modes by effectiveness. The most effective pointed out a controlling by head tilting. računalniški vid pripomoček učenje gibalne motnje computer vision tool teaching motor disorders true true false [M. Bregar] Slovenski jezik Angleški jezik Diplomsko delo/naloga 2014-09-30 12:06:30 2015-03-06 14:34:49 2022-05-18 03:06:45 2021-05-03 03:21:45 2014 0 Maribor 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004.514.66(043.2) 18541078 URN:SI:UM:DK:5SW4QGBL UNI_Bregar_Matija_2014.pdf UNI_Bregar_Matija_2014.pdf 1 AF8A3F9ED71A2673AC5E8A3C555B63F0 80b0e638548b793e68f64cebeb65fa6e7c46da304f606c29cd8894e87463f9f2 559450e1-7c0b-11eb-bb7a-00155d0001ca https://dk.um.si/Dokument.php?lang=slv&id=68067 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0