Kontrol Pergerakan Kursi Roda Berbasis Head Gesture Menggunakan CNN

Ananto, Batrisyia Zahrani (2024) Kontrol Pergerakan Kursi Roda Berbasis Head Gesture Menggunakan CNN. Diploma thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 5024201065-Undergraduate_Thesis.pdf] Text
5024201065-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 October 2026.

Download (48MB) | Request a copy

Abstract

Kuadriplegia atau tetraplegia merupakan kelumpuhan yang terjadi pada keempat anggota gerak tubuh. Karena limitasi yang mereka miliki, maka mereka memerlukan kursi roda. Mengendalikan pergerakan kursi roda bisa menjadi tantangan, terutama bagi pengguna dengan keterbatasan fisik signifikan. Diperlukan pengembangan sistem kontrol kursi roda yang dapat digunakan penderita tetraplegia. Salah satu pendekatan yang dapat dilakukan untuk mengontrol kursi roda adalah dengan melakukan ekstraksi fitur wajah dan pendeteksian head gesture dengan menggunakan mediapipe. Kemudian data head gesture yang telah diklasifikasikan akan dikirimkan oleh NUC ke sistem kontrol kursi roda. Sistem kontrol tersebutlah yang akan mengatur arah gerak kursi roda. Dengan menggunakan metodologi yang digunakan, dapat ditarik beberapa kesimpulan dari pengujian yang telah dilakukan. Model yang akan digunakan memiliki arsitektur CNN 7 layer dengan Convolutional 2D 64, 256 dan diakhiri dengan Dense 512. Jarak model yang paling tinggi akurasinya adalah 50 sentimeter. Intensitas cahaya yang paling tinggi akurasinya adalah 110 lux. Kecepatan FPS laptop penulis lebih tinggi daripada NUC yang digunakan. Rata-rata waktu delay renspons motor adalah 0,3025423729 detik dan inference time nya adalah 0,07220 detik. Rata-rata kestabilan gerak motor kursi roda untuk pendeteksian selama 2 detik adalah 8,9364 detik.
==================================================================================================================================
Quadriplegia or tetraplegia is paralysis that occurs in all four limbs. Because of their limitations, they need a wheelchair. Controlling the movement of a wheelchair can be a challenge, especially for users with significant physical limitations. It is necessary to develop a wheelchair control system that can be used by tetraplegic sufferers. One approach that can be taken to control a wheelchair is to extract facial features and detect head gesture using mediapipe. Then the classified head gesture data will be sent by the NUC to the wheelchair control system. This control system will regulate the direction of movement of the wheelchair. By using the methodology used, several conclusions can be drawn from the tests that have been carried out. The model that will be used has a 7 layer CNN architecture with Convolutional 2D 64, 256 and ending with Dense 512. The model distance with the highest accuracy is 50 centimeters. The light intensity with the highest accuracy is 110 lux. The FPS speed of the writer's laptop is higher than the NUC used. The average motor response time delay is 0.3025423729 seconds and the inference time is 0.07220 seconds. The average stability of the wheelchair motor for detection for 2 seconds is 8,9364 seconds.

Item Type: Thesis (Diploma)
Uncontrolled Keywords: Gestur Kepala, Kursi Roda, Kontrol, Tetraplegia, Control, Head Gesture, NUC, Wheelchair.
Subjects: R Medicine > R Medicine (General) > R856.2 Medical instruments and apparatus.
T Technology > T Technology (General) > T57.8 Nonlinear programming. Support vector machine. Wavelets. Hidden Markov models.
T Technology > T Technology (General) > T59.7 Human-machine systems.
T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
T Technology > TA Engineering (General). Civil engineering (General) > TA1650 Face recognition. Optical pattern recognition.
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105 Data Transmission Systems
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User: Batrisyia Zahrani Ananto
Date Deposited: 25 Jul 2024 04:45
Last Modified: 25 Jul 2024 04:45
URI: http://repository.its.ac.id/id/eprint/108772

Actions (login required)

View Item View Item