logo SBA

ETD

Digital archive of theses discussed at the University of Pisa

 

Thesis etd-10142022-101506


Thesis type
Tesi di laurea magistrale
Author
MAZZINI, MATILDE
URN
etd-10142022-101506
Thesis title
A 3D Object Detection System for Augmented Reality
Department
INGEGNERIA DELL'INFORMAZIONE
Course of study
ARTIFICIAL INTELLIGENCE AND DATA ENGINEERING
Supervisors
relatore Prof. Falchi, Fabrizio
tutor Nopper, Norbert
Keywords
  • 3d
  • Artificial
  • Artificiale
  • Augmented
  • Deep
  • Detection
  • Graphics
  • Intelligence
  • Intelligenza
  • Learning
  • Mediapipe
  • Object
  • Reality
  • Recognition
  • Tracking
  • try-on
  • Visualization
Graduation session start date
18/11/2022
Availability
Withheld
Release date
18/11/2092
Summary
The aim of this work was to investigate a state-of-the-art system that can support the
community on the visualization of virtual objects around everyday objects.
With this work, I explored 3d graphics topics and deepened my knowledge of deep learning
techniques for object detection and tracking, developing an Augmented Reality system
able to merge these two topics.
I present a Markerless 3D Augmented Reality Application for virtual accessory try-on
applications around shoes. It exploits the Mediapipe recent solution Objectron, a system
built to make 3D object recognition. The final system is a web virtual try-on application
that only uses a webcam and a browser, like Chrome, can detect the real shoe
and render 3D shoe asset on top of the shoe captured by the webcam.
File