Activity recognition from inertial sensors on a smartphone
In healthcare sector, getting to track the activity patterns of an individual is vital when providing assistance in healthcare. In this project, unobtrusive finding of a person’s activities is implemented through a smart phone. Decision of using android-based smart phone is chosen to monitor and der...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
2015
|
Subjects: | |
Online Access: | http://hdl.handle.net/10356/62788 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-62788 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-627882023-03-03T20:45:36Z Activity recognition from inertial sensors on a smartphone Chee, Kok Hao Goh Wooi Boon School of Computer Engineering Game Lab DRNTU::Engineering::Computer science and engineering In healthcare sector, getting to track the activity patterns of an individual is vital when providing assistance in healthcare. In this project, unobtrusive finding of a person’s activities is implemented through a smart phone. Decision of using android-based smart phone is chosen to monitor and derive recognition over the activities performed. There are certain issues when it first comes in raw signal, but it can be rectified via either artifact rejection which is known as filtering or by calibration. Smart phones nowadays have calibrated at most efficient way, but only up to 80-90% in terms of accuracy. An algorithm is developed to assist the activity recognition process. The effective method was to use the accelerometer sensor and gyroscope sensor’s reading to compute necessary threshold required to be categorized as activities recognizable which is walking, running, climbing staircase or in a vehicle. Test cases were conducted to retrieve results for three categories of activities namely walking, running, climbing staircase in single step and double step. These test cases are conducted over 3 individuals and average results are recorded. Evaluation results of the testing shows that the climbing staircase has the best accuracy among others. Bachelor of Engineering (Computer Science) 2015-04-29T03:10:21Z 2015-04-29T03:10:21Z 2015 2015 Final Year Project (FYP) http://hdl.handle.net/10356/62788 en Nanyang Technological University 52 p. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
DRNTU::Engineering::Computer science and engineering |
spellingShingle |
DRNTU::Engineering::Computer science and engineering Chee, Kok Hao Activity recognition from inertial sensors on a smartphone |
description |
In healthcare sector, getting to track the activity patterns of an individual is vital when providing assistance in healthcare. In this project, unobtrusive finding of a person’s activities is implemented through a smart phone. Decision of using android-based smart phone is chosen to monitor and derive recognition over the activities performed. There are certain issues when it first comes in raw signal, but it can be rectified via either artifact rejection which is known as filtering or by calibration. Smart phones nowadays have calibrated at most efficient way, but only up to 80-90% in terms of accuracy. An algorithm is developed to assist the activity recognition process. The effective method was to use the accelerometer sensor and gyroscope sensor’s reading to compute necessary threshold required to be categorized as activities recognizable which is walking, running, climbing staircase or in a vehicle. Test cases were conducted to retrieve results for three categories of activities namely walking, running, climbing staircase in single step and double step. These test cases are conducted over 3 individuals and average results are recorded. Evaluation results of the testing shows that the climbing staircase has the best accuracy among others. |
author2 |
Goh Wooi Boon |
author_facet |
Goh Wooi Boon Chee, Kok Hao |
format |
Final Year Project |
author |
Chee, Kok Hao |
author_sort |
Chee, Kok Hao |
title |
Activity recognition from inertial sensors on a smartphone |
title_short |
Activity recognition from inertial sensors on a smartphone |
title_full |
Activity recognition from inertial sensors on a smartphone |
title_fullStr |
Activity recognition from inertial sensors on a smartphone |
title_full_unstemmed |
Activity recognition from inertial sensors on a smartphone |
title_sort |
activity recognition from inertial sensors on a smartphone |
publishDate |
2015 |
url |
http://hdl.handle.net/10356/62788 |
_version_ |
1759857723593195520 |