Kiosk Interaktif: Sistem Pengendalian Touch-Free Menggunakan Gestur Tangan Berbasis MediaPipe

farhat, muhammad fathi (2025) Kiosk Interaktif: Sistem Pengendalian Touch-Free Menggunakan Gestur Tangan Berbasis MediaPipe. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5024211065-undergraduate_thesis.pdf - Accepted Version Restricted to Repository staff only Download (21MB) | Request a copy

Abstract

Pengembangan Kiosk interaktif dengan fitur virtual try-on mengubah pengalaman berbelanja di sektor retail fashion. Namun, implementasi berbasis layar sentuh menimbulkan masala
seperti kekhawatiran higienitas terutama pasca pandemi, kerusakan fisik akibat penggunaan berulang, dan keterbatasan aksesibilitas. Penelitian ini mengusulkan sistem pengendalian touchfree menggunakan gestur tangan berbasis MediaPipe untuk aplikasi virtual try-on pada Kiosk interaktif. Sistem yang dikembangkan memanfaatkan teknologi computer vision untuk mendeteksi dan melacak 21 landmark tangan, mengimplementasikan algoritma deteksi status jari berbasis sudut, dan mengembangkan dua gestur utama: pointing untuk navigasi dan selecting untuk aktivasi elemen antarmuka. Optimasi sistem mencakup adaptasi threshold berdasarkan jarak dan filtering temporal melalui riwayat gestur, menghasilkan pengenalan gestur yang akurat dan responsif dalam berbagai kondisi pencahayaan. Antarmuka pengguna dirancang khusus untuk interaksi gestur dengan implementasi elemen UI berukuran besar, umpan balik visual yang jelas, dan alur navigasi yang intuitif. Sistem ini menawarkan solusi interaksi yang higienis, tahan lama, dan aksesibel untuk aplikasi virtual try-on pada Kiosk nteraktif, memungkinkan pengguna mencoba pakaian secara virtual tanpa kontak fisik dengan permukaan Kiosk.

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

The development of interactive KIOSKs with virtual try-on features has transformed the shopping experience in the fashion retail sector. However, touchscreen-based implementations present several challenges including hygiene concerns especially post-pandemic, physical damage due to repeated usage, and accessibility limitations. This research proposes a touch-free control system using MediaPipe-based hand gestures for virtual try-on applications on interactive KIOSKs. The developed system utilizes computer vision technology to detect and track 21 hand landmarks, implements angle-based finger status detection algorithms, and develops two main gestures: pointing for navigation and selecting for interface element activation. System optimizations include distance-based threshold adaptation and temporal filtering through gesture history, resulting in accurate and responsive gesture recognition under various lighting conditions. The user interface is specifically designed for gesture interaction with the implementation of large UI elements, clear visual feedback, and intuitive navigation flow. This system offers a hygienic, durable, and accessible interaction solution for virtual try-on applications on interactive KIOSKs, allowing users to try on clothing virtually without physical contact with the KIOSK surface.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Pengendalian Touch-Free, Virtual Try-On, Gestur Tangan, MediaPipe, Kios Interaktif Touch-Free Control, Virtual Try-On, Hand Gestures, MediaPipe, Interactive Kiosk
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Muhammad Fathi Farhat
Date Deposited:	18 Jun 2025 06:13
Last Modified:	18 Jun 2025 06:13
URI:	http://repository.its.ac.id/id/eprint/119191

Actions (login required)

View Item