Monitoring of the aged and visually impaired for ambulation and activities of daily living
Nwakamma Mary Nkechinyere, Mbonu Washington Enyinna, Ossai Reginald Uche, Nwakamma Iheachoro Gerald · 2017 IEEE 3rd International Conference on Electro-Technology for National Development (NIGERCON) · 2017
The approach adopted in this context is the use of body worn accelerometer sensor to collect data to monitor the activities of daily living of an elderly person and virtually impaired. The aim is to identify the specific activities he/she has been able to perform in a day. Accelerometer sensor was used to get the data which was further analyzed in MATLAB software. The gravity accelerometer (G) result was used to create the target file for the Neural Network (NN) regression. The NN data training was done for a perfect characterization of the data before testing the data. Two individually generated data was used to validate the method to know if their activities could be accurately classified. The result shows that for both cases there was no fall with a long lye, but virtually impaired person encountered a fallfor 25 time and the aged person for 20 times with 100% correct classification.