Faculty Publications

Title

A Fall Detection System Based on a Thermopile Imaging Array and a Back Projection Algorithm

Document Type

Conference

Keywords

Back projection, Fall detection, Thermopile imaging array

Journal/Book/Conference Title

IEEE International Conference on Electro Information Technology

Volume

2020-July

First Page

60

Last Page

65

Abstract

In this paper, a fall detection system consisting of a thermopile imaging array with 80*64 pixels and a Raspberry Pi 3 has been developed. First, the thermal images captured by the hardware system are processed to eliminate fixed interferences and identify the human body. Then, the real height of the human body is estimated from the original height in the thermal images. Finally, after smoothing the fluctuation of the real height, fall events are detected according to the relative variations of the smoothed height. Our experiments show that the newly developed system and image processing algorithm can achieve much better performance on fall detection than other systems based on infrared sensors or sensor arrays.

Original Publication Date

7-1-2020

DOI of published version

10.1109/EIT48999.2020.9208276

Repository

UNI ScholarWorks, Rod Library, University of Northern Iowa

Language

en

This document is currently not available here.

Share

COinS