Article Per Year (5 Year)
p-Index From 2019 - 2024
0.23
P-Index
This Author published in this journals
All Journal Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Slamet Arifmawan
Fakultas Ilmu Komputer, Universitas Brawijaya
Computer Science & IT Control & Systems Engineering Education Electrical & Electronics Engineering Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Implementasi Algoritme Faster R-CNN untuk Sistem Pendeteksi Halangan di Luar Ruangan bagi Tunanetra Slamet Arifmawan; Fitri Utaminingrum
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 5 No 8 (2021): Agustus 2021
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

According to WHO, there are about 7 million people who recently blind every year. People who are blind or so-called visually impaired will find some limitations when doing activities, especially outdoor activities. One example of outdoor activities is walking to a certain place. There are many visually impaired people who use a white cane (a stick-like tool which has length of approximately 1.5 meters) when walking. The disadvantage of this white cane is that its range can be assumed to be only 1.5 meters, while outdoor activities require one to be more aware of the surrounding. For example, when walking on the side of the road where there are many motorized vehicles passing by. Motorized vehicles also can move fairly quickly, even in a matter of seconds they can move more than 1.5 meters so that a range of 1.5 meters is not good for outdoor activities. Starting from this problem, the author plans to build a system of assistive device for visually impaired people that adapts neural network and computer vision technology, so that the resulting range is longer than the white cane. The design of this device is a strap equipped with a processing box and a camera that is attached to the user's chest. This device has an accuracy of 91.96% using the Faster R-CNN algorithm. When using CUDA acceleration, the number of frames per second is around 1.7 fps and when not using CUDA acceleration, the number of frames per second is 0.25 fps which is about 6 time slower than using CUDA acceleration.
Co-Authors Utaminingrum, Fitri