Bu çalışma kapsamında insansız bir kara aracının kişinin el ve parmak hareketleri ile uzaktan kontrolü gerçekleştirilmiştir. Beyinden kol kaslarına iletilen ve kişinin el hareketlerini gerçekleştirmesini sağlayan Elektromiyografi (EMG) sinyalleri, kişinin koluna giydiği sekiz EMG sensör içeren bileklik vasıtası ile gerçek zamanlı olarak alınmıştır. Raspberry pi 3 gömülü sistem kartı üzerinde geliştirilen sinyal işleme, öznitelik çıkarımı ve sınıflandırma algoritmaları kullanılarak anlamlandırılmıştır. Başka bir deyişle el hareketin örüntüsü (el kapama, parmak açma, serçe parmak temas, bilek dışa bükme, vs.) ile EMG sinyal grubu arasındaki ilişkiler tanımlanmıştır. Anlamlandırılan her bir el hareketi araç için bir hareketi kontrol komutu (el kapama: araç ileri, parmak açma: araç dur, serçe parmağa temas: sola dönüş, bilek dışa bükme: sağa dönüş, vs.) olarak kullanılmıştır. Böylece insan – mobil araç etkileşim ağı kurulmuştur. Kurulan insan- mobil araç etkileşim ağı sayesinde el hareketleri ile mobil aracın gerçek zamanlı hareket kontrolü ortalama % 92 başarı ile gerçekleştirilmiştir.
In this study, remote control was carried out with the hand and finger movements of a person's unmanned ground vehicle. The electromyography (EMG) signals, which are transmitted from the brain to the arm muscles and allow the person to perform the hand movements, are taken in real time through the arm containing eight EMG sensors that the person wears in the arm. Raspberry pi 3 is understood by using signal processing, proprietary extraction and classification algorithms developed on the system card. In other words, the relationships between the hand movement sample (hand closure, finger opening, finger contact, arrow outward swelling, etc.) and the EMG signal group are defined. Each hand movement is meant to be used as a movement control command for the vehicle (hand closing: the vehicle forward, finger opening: the vehicle stop, contact with the fingers: left turn, arm swing out: right turn, etc.) Thus, a human-mobile vehicle interaction network has been established. Through the established human-mobile vehicle interaction network, hand movements and real-time control of the mobile vehicle’s movement were achieved with an average of 92% success.
In this study, remote control of an unmanned land vehicle by hand and finger movements was performed. Electromyography (EMG) signals, which are transmitted from the brain to the arm muscles and enable the person to perform hand movements, were received in real-time by a wristband containing eight EMG sensors worn on the arm. The signal processing, developed on the Raspberry pi 3 embedded system board, was recognized by using feature extraction and classification algorithms. In other words, the relationship between the pattern of hand movement (hand closure,hand opening, thumb-pinky finger touch, wrist bending, etc.) and the EMG signal group is defined. For each recognized hand gesture was used as a motion control command for the vehicle (hand closure: vehicle forward, hand opening: vehicle stop, thumb-little finger touch: left turn, wrist bend: right turn, etc.). Thus, a human - mobile vehicle interaction network was established. Thanks to the established human-mobile vehicle interaction network, real-time motion control of hand movements and the mobile vehicle were achieved with an average success rate of 92%.
Field : Fen Bilimleri ve Matematik
Journal Type : Ulusal
Relevant Articles | Author | # |
---|
Article | Author | # |
---|