EngLib USM > Ω School of Electrical & Electronic Engineering >

Arm gesture recognition system using electomyography sensor

Arm gesture recognition system using electomyography sensor / Mok Zhi Yong

Abstract Projek ini adalah mengenai pengiktirafan gerak isyarat tangan menggunakan (Electromyography) sensor EMG. Pengiktirafan isyarat boleh dilakukan dengan menggunakan dua kategori utama, salah satu adalah berdasarkan kepada sarung tangan dengan data dan yang satu lagi adalah berdasarkan komputer. Kajian ini ditubuhkan atas asas sarung tangan berdata yang boleh mengesan perubahan dalam isyarat lengan statik dan dinamik dalam masa nyata. Pada asasnya, sensor EMG telah digunakan untuk mengumpul data dan pengawal mikro (raspberry-pi) bertindak sebagai alat pemprosesan data dan klasifikasi. Alat klasifikasi berdasarkan kaedah statistik yang melibatkan Welch ujian-t. untuk pemprosesan data, kaedah statistik seperti selang keyakinan dan pengambangan telah digunakan. Isyarat tangan yang diuji dalam sistem berdasarkan gerakan bahu dan akhiran lengan. Dua sensor otot telah digunakan pada otot berbentuk delta dan bisep otot brachii. Ketepatan keseluruhan sistem ini ialah 77.5%, statik isyarat tangan mempunyai ketepatan yang lebih rendah berbanding dengan gerak isyarat tangan dinamik. _______________________________________________________________________________________________________ This project is about arm gesture recognition using EMG (electromyography) sensors. Gesture recognition can be done using two main categories, one is based on data glove and the other based on computer vision. The study is established on the basic of data sleeve which can detect changes in the static and dynamic arm gesture in real time. Basically, EMG sensor was used for data collecting and microcontroller (raspberry-pi's) acts as the data processing and classification tool. The classification tool was based on statistical method, which involved Welch’s t-test. For data processing, statistical method like confidence interval and thresholding were used. The arm gesture that were tested in the system were involved motion of shoulder and flexion of forearm. Two muscle sensor had been used on deltoid muscle and biceps brachii muscle. The overall accuracy of the system was 77.5%, static arm gesture had a lower accuracy comparing to dynamic arm gesture.

Contributor(s): Mok Zhi Yong - Author Primary Item Type: Final Year Project Identifiers: Accession Number : 875006055 Language: English Subject Keywords: gesture recognition; EMG (electromyography); computer vision First presented to the public: 6/1/2016 Original Publication Date: 7/3/2018 Previously Published By: Universiti Sains Malaysia Citation: Extents: Number of Pages - 77 License Grantor / Date Granted:   / ( View License ) Date Deposited 2018-07-03 16:09:09.943 Date Last Updated 2019-01-07 11:24:32.9118 Submitter: Mohd Jasnizam Mohd Salleh

All Versions

• Help | 
• About | 
• Privacy Policy