Accelerometry-based recognition of the placement sites of a wearable sensor

摘要: Abstract This work describes an automatic method to recognize the position of accelerometer worn on five different parts body–ankle, thigh, hip, arm and wrist–from raw data. Automatic detection body a wearable sensor would enable systems that allow users wear sensors flexibly or permit need automatically verify placement. The two-stage location algorithm works by first detecting time periods during which candidates are walking (regardless where is positioned). Then, assuming data refer walking, detects sensor. Algorithms were validated dataset substantially larger than in prior work, using leave-one-subject-out cross-validation approach. Correct placement recognition obtained for 97.4% 91.2% classified windows, respectively.