IMPLEMENTING A REAL-TIME OBJECT RECOGNITION SYSTEM BASED ON YOLOVII FOR THE BLIND

AL AZMI, ARSYA PRIMA (2026) IMPLEMENTING A REAL-TIME OBJECT RECOGNITION SYSTEM BASED ON YOLOVII FOR THE BLIND. Tugas Akhir thesis, University of Technology Yogyakarta.

[img] Text
5220411114-Arsya Prima Al Azmi_ABSTRAK.pdf
Download (136kB)

Abstract

ABSTRACT Independence in mobility poses a significant challenge for individuals with visual impairments, as they have limited ability to recognize surrounding objects visually. Conventional assistive devices, such as white canes, have limitations in detecting objects above ground level or providing specific identification that requires instant feedback. Although computer vision technology offers solutions through automatic object detection, its implementation on mobile devices is often hindered by resource efficiency constraints and high latency associated with complex models. This final project addresses this gap by optimizing the latest YOLOv11 Nano variant (YOLOv11n) architecture, integrated into the "AllScan" mobile application developed using the Flutter framework. By employing a technical approach that combines local (on-device) processing with Text-to-Speech (TTS) functionality, this research converts camera visual data into responsive speech without an internet connection. Test results demonstrate that the system can reliably recognize 30 common object classes, achieving a mean Average Precision (mAP) of 73.2% and an average confidence level exceeding 80% across various environmental conditions. The findings of this research offer a reliable, privacy-conscious assistive technology solution that can enhance safety and boost the confidence of individuals with visual impairments in their ability to interact independently with their surroundings. Keywords: Object Detection, YOLOv11n, Assistive Technology, On-Device Processing, Visual Impairment.

Item Type: Thesis (Skripsi, Tugas Akhir or Kerja Praktek) (Tugas Akhir)
Subjects: T Technology > T Technology (General)
T Technology > T Technology (General) > T201 Patents. Trademarks
Divisions: Fakultas Sains Dan Teknologi > S1 Informatika
Depositing User: Kaprodi S1 Informatika UTY
Date Deposited: 06 May 2026 03:59
Last Modified: 06 May 2026 03:59
URI: http://eprints.uty.ac.id/id/eprint/19794

Actions (login required)

View Item View Item
UTY Open Access is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.