Abstract
In this paper experiments with acceleration sensors are described for human activity recognition of a wearable device user. The use of principal component analysis and independent component analysis with wavelet transform is tested for feature generation. Recognition of the human activity is examined with multilayer perceptron classifier. Best classification results for recognition of different human motions were 83-90%, and they were achieved by utilizing independent component analysis and principal component analysis. Difference between these methods turned out to be negligible.
Original language | English |
---|---|
Pages (from-to) | 747-752 |
Number of pages | 6 |
Journal | Proceedings of the IEEE International Conference on Systems, Man and Cybernetics |
Volume | 2 |
DOIs | |
Publication status | Published - 1 Jan 2001 |
MoE publication type | A1 Journal article-refereed |
Keywords
- Accelerometer
- Independent component analysis
- Movement recognition
- Principal component analysis
- Wearable computing