Kontrol Media Player Komputer Berbasis Pose Tangan Menggunakan Convolutional Neural Network

Wiguna, I Gusti Komang Agung (2023) Kontrol Media Player Komputer Berbasis Pose Tangan Menggunakan Convolutional Neural Network. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 07211940000028-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 September 2025. Download (23MB) | Request a copy

Abstract

Tugas Akhir ini bertujuan untuk mengembangkan sistem kontrol media player komputer alternatif berbasis pose tangan menggunakan Convolutional Neural Network (CNN). Latar belakang penelitian ini yaitu kontrol media player alternatif saat ini memiliki jeda waktu lama untuk mengenali tangan dan bergantung pada warna latar belakang, yang membuatnya tidak efisien dalam segi waktu dan keakuratan kontrol. Metode penelitian ini menggunakan deteksi tangan dan 21 landmark untuk mengestimasikan pose tangan pada citra. Pendeteksian dan estimasi pose dilakukan dengan MediaPipe dan klasifikasi menggunakan CNN. Sistem ini mampu mendeteksi pose tangan dengan akurasi 97.78% dan waktu respon rata-rata 310.9 miliseconds. Sistem ini dapat mengontrol media player dengan pose tangan dari jarak 50 sampai 200 sentimeter. Kontribusi Tugas Akhir ini adalah pengembangan sistem kontrol media player alternatif untuk meningkatkan efektivitas, efisiensi, dan kecepatan respon kontrol media player komputer.
====================================================================================================================================
This Final Project aims to develop an alternative computer media player control system based on hand pose using Convolutional Neural Network (CNN). The background of this research is that the current alternative media player control has a long delay in recognizing hands and relies on the background color, making it inefficient in terms of time and control accuracy. The research method involves hand detection and 21 landmarks to estimate hand pose in images. Hand detection and pose estimation are performed using MediaPipe, and the classification is done using CNN. The system is capable of detecting hand poses with an accuracy of 97.78% and an average response time of 310.9 milliseconds. The system can control the media player based on hand poses from a distance of 50 to 200 centimeters. The contribution of this Final Project is the development of an alternative media player control system to enhance effectiveness, efficiency, and response speed in computer media player control.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Control, Estimasi pose, Kontrol, Media player, Pose estimation.
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) R Medicine > R Medicine (General) > R858 Deep Learning T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	I Gusti Komang Agung Wiguna
Date Deposited:	03 Aug 2023 04:16
Last Modified:	03 Aug 2023 04:16
URI:	http://repository.its.ac.id/id/eprint/100912

Actions (login required)

View Item