Calibration of Low-Resolution Thermal Imaging for Human Monitoring Applications

Abdallah Naser, Ahmad Lotfi (Corresponding Author), Junpei Zhong

Research output: Journal article publicationJournal articleAcademic researchpeer-review

1 Citation (Scopus)

Abstract

Thermal imaging has recently come to light to measure high human body temperature (fever) in responses to the global public health issues. This is normally achieved by very expensive high-resolution thermal cameras. Lately, there has been a new commercial low-resolution thermal sensor array (TSA) that has gained growing interest in indoor human monitoring applications due to their low-cost and human privacy-preserving claims. However, there has not been sufficient independent empirical calibration of low-resolution TSA and high-resolution images for human-centred applications. This letter provides empirical calibration of low- and high-resolution thermal imaging techniques in terms of their visible outputs, accuracy in temperature values, and stability. Besides, this letter assesses the claimed privacy-preserving feature of TSA by experimentally validating the possibility of revoking the human identity from the TSA's output. Thus, this letter aims to understand better the advantages, limitations, and future trends of using TSA in human monitoring applications.

Original languageEnglish
Article number7000904
JournalIEEE Sensors Letters
Volume6
Issue number3
DOIs
Publication statusPublished - 21 Mar 2022

Keywords

  • human centered AI
  • neural network
  • privacy-preserving
  • regression
  • sensor calibration
  • Sensor signal processing
  • thermal sensor array (TSA)

ASJC Scopus subject areas

  • Instrumentation
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Calibration of Low-Resolution Thermal Imaging for Human Monitoring Applications'. Together they form a unique fingerprint.

Cite this