Kontrol Pergerakan Kursi Roda Berbasis Hand Gesture Menggunakan Long Short-Term Memory (LSTM)

Naufal, Andrya Muhammad (2025) Kontrol Pergerakan Kursi Roda Berbasis Hand Gesture Menggunakan Long Short-Term Memory (LSTM). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5024211033-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (78MB) | Request a copy

Abstract

Penyandang disabilitas, khususnya paraplegia, menghadapi tantangan besar dalam hal mobilitas dan kemandirian. Teknologi asistif menjadi penting untuk meningkatkan kualitas hidup mereka, salah satunya dengan mengembangkan sistem kontrol kursi roda berbasis pengenalan gestur tangan. Penelitian ini bertujuan untuk mengembangkan sistem kontrol kursi roda yang dapat dikendalikan menggunakan gestur tangan dengan memanfaatkan arsitektur Long Short-Term Memory (LSTM). Sistem ini dirancang untuk mengenali lima perintah dasar gestur tangan, yaitu Maju, Mundur, Kiri, Kanan, dan Berhenti, yang dapat digunakan oleh penyandang paraplegia untuk menggerakkan kursi roda secara mandiri. Pengembangan sistem ini dilakukan dengan menggunakan teknologi visi komputer dan deep learning untuk mendeteksi dan mengklasifikasikan gestur tangan. Proses pengenalan gestur dilakukan secara real-time, dengan pengujian yang melibatkan berbagai perangkat pemrosesan, seperti laptop, Raspberry Pi 5, dan Intel NUC. Hasil pengujian menunjukkan bahwa model LSTM yang dikembangkan memiliki akurasi tinggi dalam mengklasifikasikan gestur tangan dan dapat bekerja dengan baik pada berbagai kondisi pencahayaan dan jarak. Sistem ini diharapkan dapat memberikan solusi bagi penyandang paraplegia dalam meningkatkan mobilitas mereka serta dapat diimplementasikan dalam situasi darurat untuk memberikan respons cepat terhadap ancaman atau bahaya. Dengan pengendalian yang lebih sederhana dan responsif, diharapkan kualitas hidup penyandang paraplegia dapat meningkat secara signifikan.

=================================================================================================================================

Individuals with disabilities, particularly those with paraplegia, face significant challenges in mobility and independence. Assistive technology plays a crucial role in improving their quality of life, one of which is the development of wheelchair control systems based on hand gesture recognition. This research aims to develop a wheelchair control system that can be operated using hand gestures by utilizing the Long Short-Term Memory (LSTM) architecture. The system is designed to recognize five basic hand gesture commands: Forward, Backward, Left, Right, and Stop, which can be used by individuals with paraplegia to move the wheelchair independently. The system development uses computer vision and deep learning technology to detect and classify hand gestures. The gesture recognition process is carried out in real-time, with testing involving various processing devices such as laptops, Raspberry Pi 5, and Intel NUC. The results of the testing show that the developed LSTM model has high accuracy in classifying hand gestures and performs well under varying lighting conditions and distances.
This system is expected to provide a solution for individuals with paraplegia to improve their mobility and can also be implemented in emergency situations to provide quick responses to threats or hazards. With simpler and more responsive control, it is hoped that the quality of life for individuals with paraplegia can be significantly improved.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Paraplegia, Kursi Roda, Gestur, Long Short-Term Memory (LSTM), Deep Learning, Wheelchair, Hand Gesture Recognition, Assistive Technology, Machine Learning, Computer Vision, Mobility.
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA336 Artificial Intelligence T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK4055 Electric motor T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Andrya Muhammad Naufal
Date Deposited:	18 Jun 2025 07:09
Last Modified:	18 Jun 2025 07:09
URI:	http://repository.its.ac.id/id/eprint/119195

Actions (login required)

View Item