Feature Clustering to Improve Fall Detection: A Preliminary Study
In this study, the fall detection method is carried out as stated on [1, 11]; a simple finite state machine is used to process acceleration data in sliding windows and whenever a fall-like event is found, features are extracted from this data. Using some clustering and classification algorithms described here, the event is classified as FALL or NOT_FALL. This research evaluates the performance of different proposed clustering and classification methods. It makes use of a new dataset, with data gathered by a wearable device placed on the wrist and used by several members of the research team and an emergency rescue training manikin under different fall scenarios to simulate the falls. A 10-fold cross-validation is also made to evaluate these methods on unseen data.
KeywordsFall detection Clustering Classification Wearable devices
This research has been funded by the Spanish Ministry of Science and Innovation, under project MINECO-TIN2017-84804-R.
- 2.Abbate, S., Avvenuti, M., Corsini, P., Light, J., Vecchio, A.: Monitoring of human movements for fall detection and activities recognition in elderly care using wireless sensor network: a survey. In: Wireless Sensor Networks: Application - Centric Design, p. 22. Intech (2010)Google Scholar
- 7.Hakim, A., Huq, M.S., Shanta, S., Ibrahim, B.: Smartphone based data mining for fall detection: analysis and design. Procedia Comput. Sci. 105, 46–51 (2017). http://www.sciencedirect.com/science/article/pii/S1877050917302065CrossRefGoogle Scholar
- 9.Igual, R., Medrano, C., Plaza, I.: Challenges, issues and trends in fall detection systems. Biomed. Eng. Online 12, 66 (2013). http://www.biomedical-engineering-online.com/content/12/1/66CrossRefGoogle Scholar
- 13.Meyer, D., et al.: Probability Theory Group (Formerly: E1071), TU Wien - Package ’e1071’ (2019). https://cran.r-project.org/web/packages/e1071/e1071.pdf
- 14.Purch.com: Top ten reviews for fall detection of seniors (2018). www.toptenreviews.com/health/senior-care/best-fall-detection-sensors/
- 15.R Core Team and contributors: K-means clustering in R stats package (2019). https://stat.ethz.ch/R-manual/R-devel/library/stats/html/kmeans.html
- 16.Ripley, B., Venables, W.: Functions for classification - package ‘class’ (2019). https://cran.r-project.org/web/packages/class/class.pdf
- 17.Khojasteh, S.B., Villar, J.R., de la Cal, E., González, V.M., Sedano, J., Yazg̈an, H.R.: Evaluation of a wrist-based wearable fall detection method. In: 13th International Conference on Soft Computing Models in Industrial and Environmental Applications, pp. 377–386 (2018)Google Scholar
- 18.Wu, F., Zhao, H., Zhao, Y., Zhong, H.: Development of a wearable-sensor-based fall detection system. Int. J. Telemed. Appl. (2015). https://www.hindawi.com/journals/ijta/2015/576364/
- 19.Zhang, S., Wei, Z., Nie, J., Huang, L., Wang, S., Li, Z.: A review on human activity recognition using vision-based method. J. Healthc. Eng. 2017, 31 (2017)Google Scholar
- 20.Zhang, T., Wang, J., Xu, L., Liu, P.: Fall detection by wearable sensor and one-class svm algorithm. In: Huang DS., Li K., I.G. (ed.) Intelligent Computing in Signal Processing and Pattern Recognition, Lecture Notes in Control and Information Systems, vol. 345, pp. 858–863. Springer Berlin Heidelberg (2006)Google Scholar