20.500.12556/DKUM-62192 VMESNIK MIŠICE-STROJ ZA RAZPOZNAVANJE ELEKTROMIOGRAFSKIH MERITEV MIŠIC PODLAHTI MUSCLE-COMPUTER INTERFACE FOR CLASSIFICATION OF FOREARM ELECTROMYOGRAPHIC SIGNALS V diplomskem delu smo proučili področje neinvazivnega merjenja elektromiogramov (EMG) mišic podlahti in razpoznavo gibov s pomočjo nevronskih mrež. Za zajem signalov EMG smo uporabili zapestnico Myo, iz katere smo pridobili še podatke o kotu pronacije in supinacije zapestja. Razviti programski vmesnik omogoča shranjevanje in vizualizacijo signalov EMG in razpoznava osem gibov, in sicer fleksijo, ekstenzijo, radialno deviacijo, ulnarno deviacijo, pronacijo in supinacijo zapestja ter stisnjeno pest in razširjeno dlan. S temi gibi nadzira navidezno gibanje v štirih prostostnih stopnjah. Preučili smo vpliv števila nevronov v skriti plasti nevronske mreže in vpliv števila kanalov na natančnost razpoznave gibov. V sklopu diplomske naloge smo priredili tudi odprtokodno igro, ki simulira letenje letala in za njegov nadzor uporabili gibe zapestja. muscle-computer interface, surface electromyogram, neural net, Myo armband, gesture recognition, flight simulation vmesnik mišice-stroj površinski elektromiogram nevronska mreža zapestnica Myo razpoznava gest simulacija letenja In this diploma thesis we studied non-invasive electromyogram (EMG) measurements of forearm muscles and gesture recognition with neural networks. We used Myo armband for signal capturing. Along with EMG signals we also recorded angle of forearm pronation and supination. Developed programming interface supports saving and visualisation of EMG signals and recognises eight gestures namely writs extension flexion radial deviation ulnar deviation pronation supination fist making and spread palm. With these gestures the interface controls four degrees of freedom. We have studied how the number of neurons in the hidden layer of the neural network and the number of channels affect accuracy of described gesture recognition. We also adapted an open source flight simulator game in order to control it with wrist gestures. true true false M. Jug Slovenski jezik Angleški jezik Diplomsko delo/naloga 2016-08-23 22:22:48 2016-09-06 13:44:29 2022-06-28 03:08:22 2019-07-11 12:47:00 2016 0 [Maribor 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004.5:621.391(043.2) 20144150 URN:SI:UM:DK:99XSUWZP UN_Jug_Milos_2016.pdf UN_Jug_Milos_2016.pdf 1 A6650498DC74D602B7E18760E8D61599 53aad0c882ed759c9f725dd05db9b31239cbcf512e3839d8a824c7cbddc6680e 23afe7e4-7c0d-11eb-bb7a-00155d0001ca https://dk.um.si/Dokument.php?lang=slv&id=100246 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0