Using Deep Learning with Thermal Imaging for Human Detection in Heavy Smoke Scenarios

Pei-Fen Tsai; Chia-Hung Liao; Shyan-Ming Yuan

doi:10.3390/s22145351

Using Deep Learning with Thermal Imaging for Human Detection in Heavy Smoke Scenarios

Sensors (Basel). 2022 Jul 18;22(14):5351. doi: 10.3390/s22145351.

Authors

Pei-Fen Tsai¹, Chia-Hung Liao¹, Shyan-Ming Yuan¹

Affiliation

¹ Department of Computer Science, National Yang Ming Chiao Tung University, Hsinchu 300, Taiwan.

Abstract

In this study, we propose using a thermal imaging camera (TIC) with a deep learning model as an intelligent human detection approach during emergency evacuations in a low-visibility smoky fire scenarios. We use low-wavelength infrared (LWIR) images taken by a TIC qualified with the National Fire Protection Association (NFPA) 1801 standards as input to the YOLOv4 model for real-time object detection. The model trained with a single Nvidia GeForce 2070 can achieve >95% precision for the location of people in a low-visibility smoky scenario with 30.1 frames per second (FPS). This real-time result can be reported to control centers as useful information to help provide timely rescue and provide protection to firefighters before entering dangerous smoky fire situations.

Keywords: LWIR; YOLO; convolutional neural network; evacuation in fire; firefighter protection; human detection; human rescue; infrared thermal camera; real-time object detection; smoky fire scene; thermal imaging camera.

MeSH terms

Deep Learning*
Firefighters*
Fires*
Humans
Smoke / analysis

Substances

Smoke

Grants and funding

This research was funded by the Ministry of Science and Technology of Taiwan, grant number 111-2410-H-A49-070-MY2.