Dissertations and Theses @ UNI

Availability

Open Access Dissertation

Abstract

With aging population, the importance of caring for elderly people is getting more and more attention. In this paper, a low resolution thermopile array sensor is used to develop an activity recognition system for elderly people. The sensor is composed of a 32x32 thermopile array with the corresponding 33° × 33° field of view. The outputs of the sensor are sequential images in which each pixel contains a temperature value. According to the thermopile images, the activity recognition system first determines whether the target is within the tracking area; if the target is within the tracking area, the location of the target will be detected and three kinds of activities will be identified.

Keywords- Activity Recognition, Raspberry Pi, Thermopile, Imaging Processing.

Year of Submission

2021

Degree Name

Doctor of Industrial Technology

Department

Department of Technology

First Advisor

Hong Nie, Chair

Second Advisor

Jin Zju, Co-Chair

Date Original

5-2021

Object Description

1 PDF file (vii, 70 pages)

Language

en

File Format

application/pdf

Share

COinS