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
Andika Bayhaki Al Rasyid Syah
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 Regional Convolutional Neural Network pada Sistem Pendeteksian Objek Halangan di dalam ruangan bagi Penyandang Tunanetra Andika Bayhaki Al Rasyid Syah; Fitri Utaminingrum
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 5 No 9 (2021): September 2021
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

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

Abstract

The reflection of light from objects to the eyes is used by humans to visualize an object. The role of the eye organ is very important for humans because human daily mobility requires visualization of objects that are visible to the eye. However, not everyone has optimal vision in their eyes. A person who loses his ability to see is called a blind person. The World Health Organization estimates that more than 7 million people are blind every year. Blind people usually use a walking stick. However, the use of a stick is less effective because the use of a stick can only detect objects with a relatively low distance according to the length of the stick and the use of a stick cannot classify objects that are in the way. Therefore, the researcher wants to design a system that is used as another alternative for the visually impaired for object detection aids using computer vision technology. The algorithm used is Faster R-CNN using NVIDIA Jetson Nano device and detected objects in the form of walls, doors, tables and chairs. When the system detects the presence of the object, the system will issue a notification in the form of a sound obtained from the buzzer. Based on the tests carried out, the accuracy of the Faster R-CNN algorithm training results is 95% at the number of steps 140,000. Then test the system detection time when applied to the NVIDIA Jetson Nano for 0.77 seconds for testing on images and the resulting Frame Per Second is 1.30 for real-time testing. Then the object detection accuracy is 81.25% at a distance of 2 meters and testing the accuracy of system integration with the hardware used as the system output, namely the buzzer of 100%.
Co-Authors Utaminingrum, Fitri