A Device-Orientation Independent Method for Activity Recognition外文翻译.pdfVIP

2010 International Conference on Body Sensor Networks A Device-Orientation Independent Method for Activity Recognition Surapa Thiemjarus School of Information, Computer, and Communication Technology Sirindhorn International Institute of Technology, Thammasat University (SIIT) Pathum Thani, Thailand e-mail: surapa@siit.tu.ac.th Abstract—This paper describes an orientation-independent Sensor location and orientation is an important factor that method for detecting activities of daily living based on can significantly affect the accuracy of automatic activity reference coordinate transformation. With the proposed recognition. A number of studies has been attempted to method, a classification model can be trained using data address this problem. Kunze et al [1] proposed a method for acquired during a specific sensor orientation and applied to deriving the location of an acceleration sensor placed on the other input signals regardless of the orientation of the device. user’s body based on the specific characteristics of walking The technique is validated using activity recognition motion. In [11], the effect of and a technique for dealing with experiments with four different orientations of a single tri-axial sensor displacement in a motion-based activity recognition accelerometer placed on the waist of 13 subjects performing a system are described. Lester et al [2