Deteksi Suhu Berbasis Pencitraan Termal Wajah Menggunakan Metode Novel Facial Thermal Feature Extraction

Ariesta, Salma Gilang (2022) Deteksi Suhu Berbasis Pencitraan Termal Wajah Menggunakan Metode Novel Facial Thermal Feature Extraction. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06111740000041-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Selama masa pandemi Covid-19, dilakukan upaya pencegahan penyebaran penyakit tersebut agar tidak semakin meluas. Salah satu upaya pencegahannya adalah melalui pemantauan suhu, karena suhu dapat menjadi indikator aktivitas fisiologis dan kesehatan manusia. Untuk mengurangi risiko penularan penyakit, pengukuran suhu non-kontak menjadi alternatif untuk mengukur suhu manusia tanpa kontak fisik dalam waktu yang singkat. Termografi atau pencitraan termal adalah salah satu metode pengukuran suhu non-kontak yang juga banyak digunakan pada diagnosis medis. Oleh karena itu, dalam penelitian ini dilakukan deteksi suhu berbasis pencitraan termal sebagai kontribusi pengembangan sistem pengukuran suhu non-kontak pada wajah manusia. Uji coba dilakukan dengan masukan berupa dua jenis citra, yaitu citra visual dan citra termal hasil dari pengambilan gambar oleh kamera termal FLIR C2. Pada penelitian ini dilakukan deteksi suhu terhadap empat jenis objek. Uji coba dimulai dengan melakukan deteksi facial landmark, membagi area wajah menjadi blok-blok yang disebut quadrilateral cutting, kemudian mengambil data suhu pada setiap blok, hingga menentukan suhu akhir dari objek yang terdeteksi. Dari hasil uji coba, didapatkan tingkat akurasi terbaik yaitu pada deteksi suhu objek wajah yang menggunakan masker dengan RMSE sebesar 0,67 dan MAE sebesar 0,57 derajat Celsius.
===================================================================================================================================
During the Covid-19 pandemic, efforts were made to prevent the spread of the disease from spreading further. One of the prevention efforts is through temperature monitoring, because temperature can be an indicator of physiological activity and human health. To reduce the risk of disease transmission, non-contact temperature measurement is an alternative to measure human temperature without physical contact in a short time. Thermography or thermal imaging is one of the non-contact temperature measurement methods that is also widely used in medical diagnosis. Therefore, in this study, thermal imaging based temperature detection was carried out as a contribution to the development of a non-contact temperature measurement system on the human face. The experiment was carried out with input in the form of two types of images, i.e. visual images and thermal images that were captured by the FLIR C2 thermal camera. In this study, temperature detection was carried out on four types of objects. The experiment starting with facial landmark detection, dividing the facial area into blocks called quadrilateral cutting, then taking temperature data of each block, until determining the final temperature of the detected object. From the experiment result, it was found that the best level of accuracy was in detecting the temperature of the face object using a mask with RMSE of 0,67 and MAE of 0,57 degrees
Celsius.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Deteksi suhu, non-kontak, pencitraan termal, facial landmark, quadrilateral cutting, Temperature detection, non-contact, thermal imaging, facial landmark, quadrilateral cutting
Subjects:	T Technology > TR Photography > TR267.733.M85 Multispectral imaging
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Mr. Marsudiyana -
Date Deposited:	21 Oct 2025 07:19
Last Modified:	21 Oct 2025 07:19
URI:	http://repository.its.ac.id/id/eprint/128649

Actions (login required)

View Item