Visualisasi Model Pakaian 3D Menggunakan Metode Deformasi Berbasis Data Point Cloud Dan Skeleton Pada Virtual Try-On Kiosk

Pebrianto, Surya (2026) Visualisasi Model Pakaian 3D Menggunakan Metode Deformasi Berbasis Data Point Cloud Dan Skeleton Pada Virtual Try-On Kiosk. Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 5024221047_Undergraduate_Thesis.pdf] Text
5024221047_Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (2MB) | Request a copy

Abstract

Industri ritel fashion menghadapi tantangan tingginya tingkat pengembalian produk akibat ketidaksesuaian ukuran dan ekspektasi visual. Teknologi Virtual Try-On (VTO) yang ada saat ini seringkali bergantung pada rekonstruksi avatar digital yang dapat mengurangi realisme. Penelitian ini bertujuan untuk merancang dan mengevaluasi sebuah kios VTO tiga dimensi (3D) yang secara langsung mensimulasikan pakaian pada tubuh asli pengguna secara real-time. Metodologi yang diusulkan dimulai dengan akuisisi data menggunakan kamera depth-sensing untuk menangkap representasi awan titik (point cloud) dinamis dari tubuh pengguna. Secara paralel, framework MediaPipe Holistic mengekstrak 33 landmark pose tubuh dan segmentation mask dari citra RGB. Berbeda dengan pendekatan berbasis avatar, model pakaian 3D secara langsung disesuaikan (fitted) dan dideformasi agar selaras dengan bentuk point cloud pengguna yang bergerak, dengan pergerakannya dipandu oleh landmark pose yang terdeteksi. Hasil yang diharapkan adalah sebuah purwarupa kios fungsional dengan kinerja real-time ( 30 FPS) yang divalidasi melalui pengujian kinerja dan evaluasi pengguna. Dengan menghilangkan tahap rekonstruksi avatar perantara, sistem ini diharapkan dapat memberikan pengalaman ”phygital” yang lebih otentik dan responsif, berpotensi mengurangi tingkat pengembalian produk dan meningkatkan kepuasan pelanggan di lingkungan ritel
=================================================================================================================================
The fashion retail industry faces challenges with high product return rates due to discrepancies in fit and visual expectations. Existing Virtual Try-On (VTO) technologies often rely on digital avatar reconstruction, which can reduce realism. This research aims to design and evaluate a three-dimensional (3D) VTO kiosk that directly simulates clothing onto the user’s actual body in real-time. The proposed methodology begins with data acquisition using a depth-sensing camera to capture a dynamic point cloud representation of the user’s body. In parallel, the MediaPipe Holistic framework extracts 33 body pose landmarks and a segmentation mask from the RGB stream. Departing from avatar-based approaches, the 3D clothing model is directly fitted and deformed to conform to the user’s live, moving point cloud, with its motion guided by the detected pose landmarks. The expected results are a functional kiosk prototype with real-time performance ( 30 FPS), validated through performance testing and user evaluation. By eliminating the intermediate avatar reconstruction step, this system is expected to provide a more authentic and responsive ”phygital” experience, potentially reducing product return rates and increasing customer satisfaction in a retail environment.

Item Type: Thesis (Other)
Uncontrolled Keywords: Virtual Try-On, Point Cloud, Sensor Kedalaman, MediaPipe, Teknologi Ritel, Real-time Fitting. Virtual Try-On, Point Cloud, Depth-Sensing, MediaPipe, Retail Technology, Real-time Fitting.
Subjects: Q Science > QA Mathematics > QA246.8 Gaussian
Q Science > QA Mathematics > QA401 Mathematical models.
Q Science > QA Mathematics > QA640.7 Discrete geometry
Q Science > QA Mathematics > QA75 Electronic computers. Computer science. EDP
Q Science > QA Mathematics > QA76.758 Software engineering
Q Science > QA Mathematics > QA9.58 Algorithms
T Technology > T Technology (General) > T385 Visualization--Technique
T Technology > T Technology (General) > T57.5 Data Processing
T Technology > T Technology (General) > T58.6 Management information systems
T Technology > T Technology (General) > T58.8 Productivity. Efficiency
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User: Surya Pebrianto
Date Deposited: 08 Jul 2026 01:14
Last Modified: 08 Jul 2026 01:14
URI: http://repository.its.ac.id/id/eprint/134366

Actions (login required)

View Item View Item