Görüntü İşleme Teknikleri ile İHA Tespiti
Loading...

Date
2021
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Open Access Color
OpenAIRE Downloads
OpenAIRE Views
Abstract
İnsanlık tarihinin başlangıcından günümüze kadar belirli evrelerle gelişen teknoloji, insanların ihtiyaç ve istekleri doğrultusunda, doğadan örnek alınarak ilerlemektedir. Gelişen teknolojinin insan hayatını kolaylaştırmasının yanında ülkelerin kendi öz kaynaklarını kullanabilmesi halinde savunma sanayi ile birlikte ülkelerin caydırıcılık etkisi artmaktadır. Ancak bu avantajlarının yanı sıra diğer ülke veya kötü niyetli insanların da bu teknolojilere sahip olarak doğrudan ya da dolaylı bir şekilde tehdit oluşturmaları kaçınılmazdır. Bu amaçla dünyada gelişen teknolojiyi takip edip uygulamalara katılarak daha iyisini üretmenin yanında olası tehdit içeren kullanımlara karşı önlemler almak, ülkemiz için hayati öneme sahiptir. Bu çalışmada, özellikle son dönemde gelişen İnsansız Hava Aracı ( İHA ) tespit ve tanımlanması için uygulama gerçekleştirilmiştir. Doğadan örnek alınarak, görme sisteminin bir benzeri donanımsal olarak kurulmuştur. Kurulan donanım üzerinden İHA tanımlanması ve tespiti için, son yıllarda kullanım alanı artan derin öğrenme yöntemlerinden SSD Mobilnet yöntemi ve Google alt yapısını kullanan TensorFlow kütüphanesi kullanılmıştır. Misyonu gereği belli saatlerde belli yükseklikte bulunan İHA, sürekli hareket halindedir. Bundan dolayı hızlı ve gerçek zamanlı nesne tanımlaması yapabilmek için seçilen SSD modeli bu çalışmanın ana temasını oluşturmuştur. Bu yazılım ve programların kullanımları pratik, hızlı ve ulaşılabilir donanımlarla desteklenmiştir.
Technology, which has developed in certain stages from the beginning of human history to the present, can be produced by taking samples from nature in line with the needs and desires of people. In addition to facilitating human life due to the developing technology, the ability of countries to use their own resources and deterrent effect increases with defence. However, in addition to these advantages, other countries or people with bad intentions also pose a threat in all situations and conditions, directly or indirectly, by having these technologies. For this purpose, it is of vital importance for our country to take various measures besides producing better ones by following the developing technology in the world and participating in applications. In this study, an application has been made for the detection and identification of the UAV that has developed recently. A visual system in a digital environment similar to that of nature was tried to be established. For the identification and detection of UAVs on the installed hardware, SSD Mobilnet method which is among the deep learning methods whose usage area has increased in recent years and the TensorFlow library using Google infrastructure, have been used. Due to its mission, the UAV, which is at a certain height at certain hours, is constantly in motion. For this reason, the SSD model chosen for fast and real-time object identification was the main theme of this study. These software and programs are supported with practical, fast and accessible hardware.
Technology, which has developed in certain stages from the beginning of human history to the present, can be produced by taking samples from nature in line with the needs and desires of people. In addition to facilitating human life due to the developing technology, the ability of countries to use their own resources and deterrent effect increases with defence. However, in addition to these advantages, other countries or people with bad intentions also pose a threat in all situations and conditions, directly or indirectly, by having these technologies. For this purpose, it is of vital importance for our country to take various measures besides producing better ones by following the developing technology in the world and participating in applications. In this study, an application has been made for the detection and identification of the UAV that has developed recently. A visual system in a digital environment similar to that of nature was tried to be established. For the identification and detection of UAVs on the installed hardware, SSD Mobilnet method which is among the deep learning methods whose usage area has increased in recent years and the TensorFlow library using Google infrastructure, have been used. Due to its mission, the UAV, which is at a certain height at certain hours, is constantly in motion. For this reason, the SSD model chosen for fast and real-time object identification was the main theme of this study. These software and programs are supported with practical, fast and accessible hardware.
Description
Keywords
Elektrik ve Elektronik Mühendisliği, Görüntü İşleme, Electrical and Electronics Engineering, Image Processing
Turkish CoHE Thesis Center URL
Fields of Science
Citation
WoS Q
Scopus Q
Source
Volume
Issue
Start Page
End Page
75
