作者: J. Mantyjarvi , J. Himberg , T. Seppanen
DOI: 10.1109/ICSMC.2001.973004
关键词: Wavelet 、 Principal component analysis 、 Artificial intelligence 、 Wavelet transform 、 Feature extraction 、 Sensor fusion 、 Computer science 、 Pattern recognition 、 Activity recognition 、 Independent component analysis 、 Acceleration
摘要: In this paper experiments with acceleration sensors are described for human activity recognition of a wearable device user. The use principal component analysis and independent wavelet transform is tested feature generation. Recognition examined multilayer perceptron classifier. Best classification results different motion were 83-90%, they achieved by utilizing analysis. difference between these methods turned out to be negligible.