Statistical Pattern Recognition: A Review
IEEE Transactions on Pattern Analysis and Machine Intelligence
Feature normalization and likelihood-based similarity measures for image retrieval
Pattern Recognition Letters - Special issue on image/video indexing and retrieval
The smart floor: a mechanism for natural user identification and tracking
CHI '00 Extended Abstracts on Human Factors in Computing Systems
Recognizing movements from the ground reaction force
Proceedings of the 2001 workshop on Perceptive user interfaces
The magic carpet: physical sensing for immersive environments
CHI EA '97 CHI '97 Extended Abstracts on Human Factors in Computing Systems
A practical approach to recognizing physical activities
PERVASIVE'06 Proceedings of the 4th international conference on Pervasive Computing
An automatic segmentation technique in body sensor networks based on signal energy
BodyNets '09 Proceedings of the Fourth International Conference on Body Area Networks
Hi-index | 0.00 |
Perhaps the most significant challenge in design of on-body sensors is the wearability concern. This concern requires that the size of the nodes (sensors, processing units and batteries) is minimized. Therefore, the computation and communication executed in on-body nodes must be moderated significantly. In this paper, we propose a collaborative signal processing scheme for physical movement monitoring that utilizes on-body and environmental sensors. The environmental sensor nodes perform the bulk of the signal processing and provide feedback to the on-body sensor nodes. This is due to the fact that the environmental sensor nodes have access to more powerful processing units and an unlimited energy supply. The feedback simplifies the signal processing on the on-body nodes significantly. We achieve this by performing a hierarchical classification and introducing a probabilistic measure on likelihood of possible classes for the final level of classification on on-body sensor nodes. The experimental results show the effectiveness of our method. On average the classification accuracy is reduced by 3% while the computational complexity can be scaled down by one order of magnitude compared to a global and comprehensive classification scheme.