文章基本信息

标题：Thermal Images Detection Of Covid-19 Using Convolutional Neural Network
本地全文：下载
作者：Hanaa Mohsin Ahmed ; Basma Wael Abdullah
期刊名称：Webology
印刷版ISSN：1735-188X
出版年度：2022
卷号：19
期号：1
页码：7806-7819
语种：English
出版社：University of Tehran
摘要：COVID-19 plays a central role through exceptional circumstances that have led to the disruption and suspension of many joints of life, the use of artificial intelligence to analyze thermal images helps the diagnosis of COVID-19 detection. This research hypothesized that by using the radiographic changes of COVID19 in thermal images using FLIR (Forward Looking Infrared) technology with Thermal Camera to detect peoples’ infection of Covid-19 by the use of intelligence systems can extract certain graphical elements related to COVID19 and offer a clinical diagnosis before pathogenic testing; therefore, vital time is saved for disease prevention. Using real dataset which collected by the author for the period 1st August to 15th September which contains 50 images for healthy peoples besides 50 patients whom infected with Covid-19. CNN (convolutional neural network) model inspired by the Xception architecture was presented for the diagnosis of patients infected with coronavirus pneumonia. The suggested technique achieved 0.92 as an average training precision. Ultimately, these findings revealed that Deep Learning (DL) model can enhance early diagnosis, treatment and isolation; therefore, it might help to manage the crisis.
关键词：Convolutional Neural Network;COVID-19;Deep Learning;Detection Thermal images