• English
    • Norsk
  • English 
    • English
    • Norsk
  • Administration
View Item 
  •   Home
  • Det matematisk-naturvitenskapelige fakultet
  • Institutt for informatikk
  • Institutt for informatikk
  • View Item
  •   Home
  • Det matematisk-naturvitenskapelige fakultet
  • Institutt for informatikk
  • Institutt for informatikk
  • View Item
JavaScript is disabled for your browser. Some features of this site may not work without it.

Using Skeleton Information for Human Identification for Elderly Care and Alarm System with the XBox One Kinect Sensor

Qi, Lexu
Master thesis
View/Open
Master-Thesis-2017-.pdf (1.383Mb)
Year
2017
Permanent link
http://urn.nb.no/URN:NBN:no-61913

Metadata
Show metadata
Appears in the following Collection
  • Institutt for informatikk [3603]
Abstract
This project explores the possibility of human identification for the Care and Alarm System by using human skeleton information captured by the Xbox One Kinect sensor. This project experiment three classifiers and three main different kinds of skeleton information. These three classifiers are artificial neural networks (ANNS), K nearest neighbours (KNNs) and support vector machine (SVM); three main kinds of skeleton information include absolute full-body skeleton information, absolute partly body skeleton information and proportional full-body skeleton information. According to the result, ANNs perform the best classification with an accuracy rate of 99.50% by using absolute full-body skeleton information. KNN generally has a high classification accuracy higher than 90% by using all kinds of skeleton information. SVM is not a good choice since it only results in bad classification results. Regarding the skeleton information, Absolute full-body produces a better classification result than proportional full-body information and absolute partly body information. Additionally, for absolute partly body information, torso information leads to a better result than bottom part. Generally speaking, skeleton information is applicable for human identification and can help generate good classification result.
 
Responsible for this website 
University of Oslo Library


Contact Us 
duo-hjelp@ub.uio.no


Privacy policy
 

 

For students / employeesSubmit master thesisAccess to restricted material

Browse

All of DUOCommunities & CollectionsBy Issue DateAuthorsTitlesThis CollectionBy Issue DateAuthorsTitles

For library staff

Login
RSS Feeds
 
Responsible for this website 
University of Oslo Library


Contact Us 
duo-hjelp@ub.uio.no


Privacy policy