Engineering, Technology & Applied Science Research Vol. 13, No.

1, 2023, 9961-9967 9961

A Real-Time Analytic Face Thermal

Recognition System Integrated with Email
Notification
Ranjit Singh Sarban Singh
Centre for Telecommunication Research & Innovation (CeTRI), Fakulti Kejuruteraan Elektronik dan
Kejuruteraan Komputer (FKEKK), Universiti Teknikal Malaysia Melaka (UTeM), Malaysia
ranjit.singh@utem.edu.my
(corresponding author)

T. Joseph Sahaya Anand

Sustainable and Responsive Manufacturing, Fakulti Teknologi Kejuruteraan Mekanikal dan Pembuatan,
Universiti Teknikal Malaysia Melaka (UTeM), Malaysia
anand@utem.edu.my

Siti Aisyah Anas

Centre for Telecommunication Research & Innovation (CeTRI), Fakulti Kejuruteraan Elektronik dan
Kejuruteraan Komputer (FKEKK), Universiti Teknikal Malaysia Melaka (UTeM), Malaysia
aisyah@utem.edu.my

Biswaranjan Acharya
Department of Computer Engineering-AI & BD, Marwadi University, India
biswaacharya@ieee.org

Received: 21 October 2022 | Revised: 16 November 2022 | Accepted: 19 November 2022

ABSTRACT
COVID-19 is a contagious disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-
CoV-2). The disease has spread worldwide, leading to an ongoing pandemic. The most common symptom
of COVID-19 is fever which can be detected using various manual screening techniques that have the risk
of exposing the personnel. Since the virus has globally spread, a reliable system to detect COVID-19-
infected people, especially before entering any premises and buildings, is in high demand. The most
common symptom that can be detected is fever, even though people with fever might not have COVID-19.
Thus, a real-time analytic face thermal recognition system integrated with email notification that has the
capability to scan the person’s temperature and simultaneously analyze the measured temperature with
the recorded/stored information/data is presented in this paper. The proposed system is also able to send
an email notification to the relevant authorities during the real-time analytical process. Besides that, this
information is also recorded in the system database for continuous monitoring of the respective person’s
health status. The development of the proposed system is integrated with a Thermal Module AMG8833, Pi
camera, and Raspberry Pi Zero Wireless. The proposed system has been tested and the captured results
successfully accomplished the development objectives.

Keywords-Covid-19; face recognition; temperature; real-time analysis; email notification

I. INTRODUCTION Procedure (SOP), a real-time analytic face thermal recognition

system integrated with email notification and having the
There is a high demand in face thermal recognition systems capability to simultaneously analyze the measured person’s
due to the COVID-19 pandemic. There are many off-the-shelf temperature and update recorded/stored information/data is
face thermal recognition systems available in the market, presented in this paper. Prior to the development of the
everyone with its abilities and advantages. To enable effective proposed system, off-the-shelf systems related to detecting
monitoring which also complies with the Standard Operating possible symptom of COVID-19 are studied. The studied

www.etasr.com Singh et al.: A Real-Time Analytic Face Thermal Recognition System Integrated with Email …
 Engineering, Technology & Applied Science Research Vol. 13, No. 1, 2023, 9961-9967 9962

systems investigate the aspect of simplicity in terms of product temperature reading and compares it with the preset
development as well as quick and fast delivering the analyzed temperature value before sending an email notification if the
and processed information to the relevant authorities. Although temperature is above the preset value. The email notification
the developed system is not able to prevent the COVID-19 notifies the relevant authorities about the person’s condition, or
pandemic, it can assist in controlling it more effectively and records the person absence from work. Phyton programming
efficiently. Generally, infrared and wireless thermometers are language was used to develop the proposed research project.
deployed as general medical tools, especially at the entries and Thonny application was selected as Python IDE because it is
exits of buildings. The infrared and wireless thermometers free and open-source and it provides significant features such
generally use body temperature sensing [1]. Designing and as code inspection and debugging. The standard libraries used
developing systems able to assist the early detection of any in this project are OpenCV and DLIB. The algorithm
fever-related illness so that it can be further examined it is development begins with facial detection and face recognition.
CoVID-19 or something else is a scientific topic of major Then, the AMG8833 Thermal Camera Sensor for facial
interest. temperature measuring, recording, and comparing is activated.
In [2], fever is mentioned as the most common symptom of The proposed project continuously processes the
patients with COVID-19. The off-the-shelf studied systems [3- recorded/stored information/data and automatically updates the
5] use the Pi Camera to visualize the thermal data, where the stored data into the developed system. Also, the developed and
temperature values are mapped into a color gradient, ranging deployed system tends to reduce the risk of infection of manual
from blue to red. The blue color is mapped to value 0 screening.
representing the lowest temperature, while red color is mapped
to value 1023 representing the highest temperature. A real II. RESEARCH METHODOLOGY
image of the respective person is blended with the saved image The proposed development methodology is divided into
to produce the read thermal temperature value. The produced two phases. Phase one focuses on the hardware development
thermal temperature value is displayed in real-time and an which consists of 1) the integration of the Pi Camera to
email is sent to the person in charge at the monitoring station if perform the facial detection and recognition and 2) the
any abnormality is detected. In [6], an IoT smart health integration of the thermal camera sensor to measure facial
monitoring system that analyzes the human body temperature, temperature. The integration of the Pi Camera is conducted to
pulse and SpO2 has been developed and integrated into a perform the face detection and recognition via existing images
mobile application for COVID-19 detection. With the that have been preloaded into the Raspberry Pi Zero Wireless
integrated system, the patient can immediately seek medical SD-Card storage. Upon the success of the face detection and
attention if a physician is not available. In [7], a smart helmet is recognition, the thermal camera sensor is activated to measure
developed to innovatively detect and monitor COVID-19 using the human facial temperature. Then, the recognized person and
a thermal imaging system attached onto the smart helmet. The its measured temperature details are stored into the Raspberry
smart helmet detects high body temperature and sends the Pi Zero Wireless SD-Card. The second phase focuses on the
measured value into a mobile application, assisting the early embedded software development for 1) facial detection and
detection of COVID-19. Thermal face recognition systems recognition, 2) measurement of the facial temperature, and 3)
have been proven to be effective in detecting possible COVID- configuration of the email notification. The facial detection and
19 affected persons besides from been used to authorize the recognition perform real-time face recognition based on the
person’s authentication [5]. Authors in [8] utilized the smart preloaded images. Upon detection of a human face, the thermal
home concept to detect if a visiting person is infected with camera sensor measures the respective human facial’s
COVID-19. Infrared technology was used to detect the temperature and stores it into the SD-Card. While the
person’s temperature and if it was high, an integrated thermal recognized face and human body temperature information is
face camera was used to scan and monitor the person’s recorded, the process of comparing the measured temperature
condition at the door. Authors in [9] mentioned that CT-scans with the preset temperature value is performed to detect normal
and X-ray images can also be used for COVID-19 early and high temperature. If the measured temperature is above the
detection. The paper also explains that the use of deep learning preset temperature value, an email notification is delivered to
models [10, 11] could give more precise COVID-19 early the relevant authorities (security, managers, internal clinic,
detection, but systems with deep learning models is more COVID-19 screening team, etc.).
expensive. Thermal detection is a popular method to screen
people during the outburst of the COVID-19 pandemic. The A. IoT Face Thermal Recognition System Integration
infrared thermal scanning method is also widely used at many  Pi Camera Integration
places as an early precaution step. Such systems are often able
to make buzzer sounds when a person with high temperature is In the first phase, the Pi Camera performs facial detection
detected during the scanning process. and recognition. The integration is shown in Figure 1. The Pi
Camera is connected to the Raspberry Pi Zero Wireless via the
After reviewing some of the developed systems that use Camera Serial Interface (CSI) connector. The integrated Pi
facial thermal measuring for early COVID-19 detection, and to Camera has the capability to capture an 8 Megapixel real-time
the best of our knowledge, features such as real-time image which will be used to perform the facial detection and
identification/authentication and email notification are not recognition with the images stored in the Raspberry Pi Zero
applicable. Hence, the current research project implements Wireless SD-Card Storage.
real-time facial identification/authentication, then analyzes the

www.etasr.com Singh et al.: A Real-Time Analytic Face Thermal Recognition System Integrated with Email …
 Engineering, Technology & Applied Science Research Vol. 13, No. 1, 2023, 9961-9967 9963

 Integration of the AMG8833 Thermal Camera Sensor

This section describes the AMG8833 Thermal Camera
Sensor connectivity with the Raspberry Pi Zero Wireless as
shown in Figure 2. The General-Purpose Input/Output (GPIO)
port 2 and port 3 are respectively connected to the Serial
Information (SDA) and Serial Clock Line (SCL). The power
source (VCC) is connected to 3.3V at port 1 and the ground is
connected to port 9. The AMG8833 Thermal Camera Sensor is
activated when the facial detection and recognition is
completed.

Fig. 1. Pi Camera and Raspberry Pi Zero Wireless connection.

Fig. 3. Embedded face detection and recognition algorithm.

Fig. 2. AMG8833 Thermal Camera Sensor connection with the Raspberry

Pi Zero Wireless.

B. Software Design and Development

 Embedded facial detection and recognition algorithm
To perform the facial detection and recognition tasks, a
suitable algorithm has been implemented (Figure 3). When the
integrated Pi Camera is active, it will acquire a real-time image
which will be compared with a preloaded image. Upon
confirming the acquired real-time image, the embedded
software algorithm will perform facial recognition to confirm
the person’s identity and will provide the recognized person’s
name. Once the name has been confirmed, it will be recorded
into the attendance list.
Fig. 4. Embedded AMG8833 thermal sensor algorithm.

www.etasr.com Singh et al.: A Real-Time Analytic Face Thermal Recognition System Integrated with Email …
 Engineering, Technology & Applied Science Research Vol. 13, No. 1, 2023, 9961-9967 9964

 Embedded algorithm for the AMG8833 thermal camera

sensor and email notification
The AMG8833 thermal camera sensor is integrated to
measure the real-time human temperature. The AMG8833
sensor reads a real-time temperature value when a person’s
face approaches the camera and the sensor. Hence, when the
face is recognized, the AMG8833 thermal camera sensor is
activated to measure the person’s facial temperature and match
the measured temperature with the person’s name that is stored
in the attendance list. If the person’s name is matched, the
name and measured temperature will be displayed at the
integrated Liquid Crystal Display (LCD). Otherwise, the
embedded software continues to perform the person’s name Fig. 6. Placement of integrated Pi Camera, AMG8833 Thermal Camera
search to record the measured temperature value. Also, during Sensor, and LCD - IoT face thermal recognition system.
the temperature measuring, the embedded software algorithm
for the AMG8833 thermal camera sensor categorizes the
measured temperature into two classes, i.e. "35°C ≤ NORMAL
≤ 37°C" or "37°C < HIGH ≤ 40°C". If the measured
temperature is Normal, its value is displayed on the LCD,
otherwise an email notification will be sent to the authorities to
inform about the person’s condition as depicted in Figures 4
and 5.

Fig. 7. Illustration of face recognition in a rectangular box.

Another reason of placing the AMG8833 sensor and the Pi

Camera side by side is because this way the AMG8833 thermal
camera sensor picture size of 480×480 pixels is centered at the
Pi Camera picture size of 640×480 pixels as shown in Figure 7.
This allows the AMG8833 Thermal Camera Sensor to
effectively scan the center of the forehead for temperature
measuring as can be seen in Figure 7. Figure 8 shows the
complete integration of Pi Camera, AMG8833 thermal camera
sensor, and LCD into the Raspberry Pi Zero Wireless.

Fig. 5. Embedded real-time email processing and notification algorithm.

III. RESULTS AND DISCUSSION

A. Hardware System Design and Development
Figure 6 shows the arrangement of AMG8833 thermal
camera sensor, Pi Camera, and LCD for the hardware system Fig. 8. Integration of the Pi Camera, AMG8833 Thermal Camera Sensor,
design and development. The AMG8833 and the Pi Camera are and LCD in the Raspberry Pi Zero Wireless.
placed side by side to allow simultaneous facial detection and
recognition and facial temperature measurement upon B. Validation of the Developed Software
authentication. Details such as name and temperature of the This section validates the developed software algorithm
scanned face are displayed on the integrated LCD. which integrates the Pi Camera and AMG8833 Thermal
Camera Sensor in Raspberry Pi Zero Wireless for face

www.etasr.com Singh et al.: A Real-Time Analytic Face Thermal Recognition System Integrated with Email …
 Engineering, Technology & Applied Science Research Vol. 13, No. 1, 2023, 9961-9967 9965

detection, face recognition, and face temperature reading and The AMG8833 thermal camera sensor displays a real-time
measurement. To perform face detection and recognition a temperature value when a person’s face approaches the Pi
suitable algorithm (Figure 4) was implemented. The system Camera. Hence, once the face is recognized, the AMG8833
starts with the embedded algorithm for face detection and thermal camera sensor is activated to measure the person’s
recognition. When the integrated Pi Camera is activated it will facial temperature and match the measured temperature with
acquire a real-time image, which will be compared with the the person’s name before storing the information into the SD-
stored image in the SD-Card. Upon confirming the acquired card and records the person’s attendance. Figure 9 also shows
real-time image, the embedded software algorithm will perform the Phyton shell output which validates the information that has
face recognition to confirm the person’s identity. Once the been stored into the SD-card. The output of the Phyton shell is
person’s face is recognized, the AMG8833 thermal camera also shown on the LCD as shown in Figure 10.
sensor will be activated to measure the person’s facial
temperature which will be evaluated based on the two C. Conditions of the Face Thermal Recognition System
conditions shown in Figures 4 and 5. These conditions will be When the face detection and recognition are successfully
matched while real-time facial temperature is measured and the validated, the program reads the temperature of the person. As
information is stored into the Raspberry Pi Zero Wireless SD- mentioned above, fever is the most common symptom of
Card. The person attendance status is also updated upon COVID-19. This stage is important to validate the system’s
completing the temperature measuring process. operation. If the system is unable to distinguish between a low-
risk and a high-risk individual, this system failure will cause
Figure 9 shows the captured result from the Python shell in possible spread of the COVID-19 disease. According to the
the Raspberry Pi Zero Wireless when the program starts to World Health Organization (WHO), normal human body
operate. After the system successfully acquires the preloaded temperature ranges from 35.0°C to 37.1°C as shown in Figure
image, the person’s face detection and recognition are 4 and people with this range of facial or body temperature are
executed. During the execution, the person’s name is validated considered healthy and fit to work. A person with a fever is
and the proposed system continues to measure the facial defined as someone who has a body temperature greater than
temperature by using the AMG8833 sensor and stores the 37.1°C (Figure 5) [12-14]. So, the temperature scanned values
measured temperature value into the SD-card. Figure 9 shows are classified as Normal and High for the respective
the name and temperature of the person after scanning, using temperatures as shown in Table I.
the Pi Camera and the AMG8833 thermal camera sensor.
TABLE I. CONDITIONS AND TEMPERATURES
Conditions Temperature (˚C)
Normal 35.0 – 37.0
High 37.1 – 40.0

 Condition 1
If the measured temperature is between 35°C and 37°C
(Normal), the temperature value will be displayed on the LCD
when the person’s name is matched (Figure 11). Once the
person’s name has been confirmed and the AMG8833 thermal
camera sensor has successfully read the normal temperature of
the user, the name and the temperature will be recorded into the
attendance list accordingly. Similarly, the person’s data are also
recorded into the system’s database to provide real-time health
Fig. 9. Name and captured temperature– Phyton shell - Raspberry Pi Zero monitoring update of the person. This attendance record can
Wireless. also be sent to relevant authorities as shown in Figure 11.
 Condition 2
If the measured temperature is between 37oC and 40oC,
then the person is not allowed to enter the premise or building.
The person's information/data are sent through an email
notification to the competent authorities. Also, during the
temperature scanning, the status "Temperature High" is
displayed on the LCD and at the Python shell as shown in
Figure 13. Figure 14 shows an email notification containing
information about the scanned person’s temperature.
D. Summary
The presented results successfully validated the proposed
Fig. 10. Person’s name and measured temperature are displayed at the real-time analytic face thermal recognition system integrated
integrated LCD screen.
with email notification. The system did successfully perform
the face detection and recognition tasks, measured the facial

www.etasr.com Singh et al.: A Real-Time Analytic Face Thermal Recognition System Integrated with Email …
 Engineering, Technology & Applied Science Research Vol. 13, No. 1, 2023, 9961-9967 9966

temperature of the scanned person, and sent an email IV. CONCLUSION

notification when required. Therefore, the proposed system's A real-time analytic face thermal recognition system
design and integration method can be implemented into integrated with email notification was developed and deployed
premises or buildings to assist the early detection of any fever- in order to assist the early detection of any fever-related illness,
related illness. especially COVID-19. Due to the severity of the COVID-19
disease, the proposed system has integrated some features that
may assist the authorities and keep the premises or buildings
safe. Also, the integrated hardware and software successfully
validated the objectives of the real-time analytic face thermal
recognition system integrated with email notification. In terms
of system deployment, the system can be placed in any open
space, where the respective person is needed to perform
temperature scanning before entering a building or premise.
ACKNOWLEDGMENT
The authors gratefully acknowledge the support of the
Centre of Telecommunication Research & Innovation (CeTRI),
Fakulti Kejuruteraan Elektronik dan Kejuruteraan Komputer
(FKEKK), Universiti Teknikal Malaysia Melaka and the
Ministry of Higher Education, Malaysia.
REFERENCES
Fig. 11. Normal temperature condition. [1] R. Saher and M. Anjum, "Role of technology in COVID-19 pandemic,"
Researches and Applications of Artificial Intelligence to Mitigate
Pandemics, vol. 2021, pp. 109–138, 2021, https://doi.org/10.1016/B978-
0-323-90959-4.00005-5.
[2] A. Barnawi, P. Chhikara, R. Tekchandani, N. Kumar, and B. Alzahrani,
"Artificial intelligence-enabled Internet of Things-based system for
COVID-19 screening using aerial thermal imaging," Future Generations
Computer Systems, vol. 124, pp. 119–132, Nov. 2021, https://doi.org/
10.1016/j.future.2021.05.019.
[3] R. E. Tabiah, M. I. Solihin, and A. Machmudah, "Developing Affordable
IoT-based Body Temperature Screening," Journal of Sustainable
Engineering and Built Environment, pp. 1–5, 2020.
Fig. 12. File CSV- individual name and temperature - attendance list. [4] V. Manuel Ionescu and F. Magda Enescu, "Low cost thermal sensor
array for wide area monitoring," in 2020 12th International Conference
on Electronics, Computers and Artificial Intelligence (ECAI), Bucharest,
Romania, Jun. 2020, https://doi.org/10.1109/ECAI50035.2020.9223193.
[5] M. Krišto and M. Ivasic-Kos, "An overview of thermal face recognition
methods," in 2018 41st International Convention on Information and
Communication Technology, Electronics and Microelectronics
(MIPRO), Opatija, Croatia, Feb. 2018, pp. 1098–1103, https://doi.org/
10.23919/MIPRO.2018.8400200.
[6] M. M. Khan, S. Mehnaz, A. Shaha, M. Nayem, and S. Bourouis, "IoT-
Based Smart Health Monitoring System for COVID-19 Patients,"
Computational and Mathematical Methods in Medicine, vol. 2021, Nov.
2021, Art. no. e8591036, https://doi.org/10.1155/2021/8591036.
[7] Assoc. Prof. Dr. M. Abdulrazaq, H. Zuhriyah, S. Al-Zubaidi, S. Karim,
R. Ramli, and E. Yusuf, "Novel Covid-19 Detection and Diagnosis
Fig. 13. High temperature condition. System Using Iot Based Smart Helmet," International Journal of
Psychosocial Rehabilitation, vol. 24, no. 7, pp. 2296–2303, Mar. 2020.
[8] A. J. Ismael, H. A. Rahim, R. Ghazali, and K. N. Mezher, "Smart Home
and Thermal Imaging Technology for COVID 19 Detection Using the
Internet of Things (IoT)," ELEKTRIKA- Journal of Electrical
Engineering, vol. 20, no. 2–2, pp. 87–92, Sep. 2021.
[9] N. Kumar, A. Hashmi, M. Gupta, and A. Kundu, "Automatic Diagnosis
of Covid-19 Related Pneumonia from CXR and CT-Scan Images,"
Engineering, Technology & Applied Science Research, vol. 12, no. 1,
pp. 7993–7997, Feb. 2022, https://doi.org/10.48084/etasr.4613.
[10] H. A. Owida, A. Al-Ghraibah, and M. Altayeb, "Classification of Chest
X-Ray Images using Wavelet and MFCC Features and Support Vector
Machine Classifier," Engineering, Technology & Applied Science
Fig. 14. Email notification. Research, vol. 11, no. 4, pp. 7296–7301, Aug. 2021, https://doi.org/
10.48084/etasr.4123.

www.etasr.com Singh et al.: A Real-Time Analytic Face Thermal Recognition System Integrated with Email …
 Engineering, Technology & Applied Science Research Vol. 13, No. 1, 2023, 9961-9967 9967
[11] S. A. A. Biabani and N. A. Tayyib, "A Review on the Use of Machine
Learning Against the Covid-19 Pandemic," Engineering, Technology &
Applied Science Research, vol. 12, no. 1, pp. 8039–8044, Feb. 2022,
https://doi.org/10.48084/etasr.4628.
[12] X. Fu, "Design of Facial Recognition System Based on Visual
Communication Effect," Computational Intelligence and Neuroscience,
vol. 2021, Dec. 2021, Art. no. e1539596, https://doi.org/10.1155/
2021/1539596.
[13] K. H. Rahouma and A. Z. Mahfouz, "Design and Implementation of a
Face Recognition System Based on API mobile vision and Normalized
Features of Still Images," Procedia Computer Science, vol. 194, pp. 32–
44, Jan. 2021, https://doi.org/10.1016/j.procs.2021.10.057.
[14] V. K. Sharma, "Designing of Face Recognition System," in 2019
International Conference on Intelligent Computing and Control Systems
(ICCS), Madurai, India, Feb. 2019, pp. 459–461, https://doi.org/
10.1109/ICCS45141.2019.9065373.

www.etasr.com Singh et al.: A Real-Time Analytic Face Thermal Recognition System Integrated with Email …