Deteksi Suhu Melalui Citra Termal Wajah Menggunakan Deep Learning

Aryasatya, Dion Devara (2021) Deteksi Suhu Melalui Citra Termal Wajah Menggunakan Deep Learning. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

Text 05111740000080-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (2MB) | Request a copy

Abstract

Dalam masa pandemi kasus penularan virus CORONA masih tetap bertambah dari hari ke hari. Salah satu gejala yang umum dialami oleh pasien COVID-19 adalah demam. Hal yang umum dilakukan untuk mengukur suhu di masa pandemi adalah menggunakan termometer non kontak. Deep Learning bisa digunakan untuk mendeteksi wajah dan membantu mendeteksi suhu maksimal wajah dari gambar termal.
Tujuan Tugas Akhir ini adalah membuat aplikasi pendeteksi suhu pada citra termal menggunakan pendekatan Deep Learning. Dalam Tugas Akhir ini dilatih sebuah model deteksi SSD-MobileNet untuk mendeteksi area wajah dari citra termal. Setelah terdeteksi, data suhu diekstrak dari area wajah tersebut.
Dalam pelaksanaan Tugas akhir ini digunakan dataset citra termal Tuft Face Database, IRDatabase, dan citra termal yang diambil menggunakan Flir One. Dari hasil uji coba didapatkan hasil mean average precision deteksi wajah sebesar 0,95 dengan threshold dari evaluasi model untuk IoU 0,75 sebesar 0,95 dan mean absolute error deteksi suhu sebesar 1,51.
================================================================================================
During the pandemic, cases of transmission of the CORONA virus are still increasing day by day. One of the common symptoms experienced by COVID-19 patients is fever. The common thing to do to measure temperature during a pandemic is to use a non-contact thermometer. Deep Learning can be used to detect faces and help detect the maximum face temperature from thermal images.
The purpose of this final project is to create a temperature detection application on thermal images using a Deep Learning approach. In this final project, an SSD-MobileNet detection model is trained to detect facial areas from thermal images. Once detected, temperature data is extracted from that facial area.
In the implementation of this final project, the Tuft Face Database thermal image dataset, IRDatabase, and thermal images taken using Flir One are used. From the test results, the mean average precision of face detection is 0,95 with the threshold of the model evaluation for IoU 0,75 of 0,95 and the mean absolute error of temperature detection of 1,51.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Deteksi Suhu, Deep Learning, SSD-Mobilenet, Citra termal Temperature Detection, Deep Learning, SSD-Mobilenet, Thermal Imaging
Subjects:	Q Science > QC Physics > QC271 Temperature measurements R Medicine > R Medicine (General) > R858 Deep Learning
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Dion Devara Aryasatya
Date Deposited:	15 Aug 2021 03:48
Last Modified:	15 Aug 2021 03:48
URI:	http://repository.its.ac.id/id/eprint/86692

Actions (login required)

View Item