A Study on Automatic Analysis Method of Human Behavior Using K-Mean Clustering of Smartphone Acceleration Sensor

스마트폰 가속도 센서의 K-평균 클러스터링을 이용한 사람행동 자동분석 방법에 대한 연구

  • Park, Jong-Kun (Division of Convergence Computer and Media, Mokwon University) ;
  • Song, Teuk-Seob (Division of Convergence Computer and Media, Mokwon University)
  • 박종권 (목원대학교 융합컴퓨터 미디어학부) ;
  • 송특섭 (목원대학교 융합컴퓨터 미디어학부)
  • Published : 2019.05.23


Smartphones have various sensors built in. In particular, acceleration sensors are used to analyze human behavior because they can detect movement of objects. Previous studies have analyzed the behavior of people by analyzing the magnitude of acceleration sensor values. In this study, we proposed a method of detecting the motion by applying the K-average of the acceleration sensor value built in the smartphone. We proposed a method of recognizing walking and running, which is basic human behavior, by applying K-average of acceleration sensor value of smartphone.


Supported by : 한국연구재단