Abstract— The stronger the trend towards nuclear families gets, the more elderly live without younger people who take care of them. It makes the elderly feel isolated and less active. The decrease of the activity weakens their muscle strength. There are many factors to prevent them from independently living. We need a method to confirm those elderly keep good health conditions enough to spend daily lives independently. This paper proposes a method to detect abnormal features from the elderly life logs. We use life logs indicating where he is and what he touches. Abnormal symptoms can be detected with the comparison of daily life logs with the derived features. Through an experiment, the proposed method has detected 4 features of dull arm and 2 features of dull leg. One feature means attributes common to objects which he does not touch when he feels his dominant arm dull. Another means attributes common to objects which he touches with his nondominant arm when he feels his dominant arm dull. In addition, the proposed method has detected that his steps increase and his length of stride is shorten in cleaning when he feels his leg dull.
Index Terms— The elderly abnormal features, RFID, life log.
Naotaka Sato is a graduate school student in Rittsumeikan University, Kusatsu, Shiga, Japan (e-mail: naotake@de.is.ritsumei.ac.jp).
The authors are with Information and science college in Ritsumeikan University, Kusatsu, Shiga, Japan (e-mail: harada@cs.ritsumei.ac.jp)
[PDF]
Cite: Naotaka Sato, Fumiko Harada, and Hiromitsu Shimakawa, " Detecting Dull Features from Action Logs of Elderly," International Journal of Information and Electronics Engineering vol. 3, no. 5, pp. 457-460, 2013.