Open AccessDeteksi Sampah pada Real-time Video Menggunakan Metode Faster R-CNN
Author(s) -
Maryam Rahman,
Bambang Bambang
Publication year - 2021
Publication title -
applied technology and computing science journal
Language(s) - English
Resource type - Journals
eISSN - 2621-4474
pISSN - 2621-4458
DOI - 10.33086/atcsj.v3i2.1846
Subject(s) - frame (networking) , computer science , garbage , artificial intelligence , computer vision , ignorance , object (grammar) , frame rate , real time computing , computer graphics (images) , telecommunications , philosophy , epistemology , programming language
Garbage is a never-ending problem in human life. Many of the problems caused by waste actually stem from human ignorance of the environment. Several solutions have been proposed to solve and avoid problems from the waste, one of which is making waste detection that can be applied directly to certain devices. This study aims to apply an object detection method in the form of Faster R-CNN to detect and classify at a speed that allows computers to detect trash objects directly through real-time video. The test results in this study indicate the method used can detect trash objects in 100 images with an accuracy value of 74%, and to detect real-time video with video frame rates in the range of 1 frame per second (fps).