EEG Data Compression using Truncated Singular Value Decomposition for Remote Driver Status Monitoring

Advancements in wireless body sensor technology have enabled continuous recording of Electroencephalogram (EEG) data for remote monitoring. However, a significant amount of data introduced due to the continuous data recording over time has become a challenge for energy constraint sensor nodes to tra...

Full description

Saved in:
Bibliographic Details
Main Authors: Alam, M.K., Aziz, A.A., Latif, S.A., Awang, A.
Format: Conference or Workshop Item
Published: Institute of Electrical and Electronics Engineers Inc. 2019
Online Access:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85075640107&doi=10.1109%2fSCORED.2019.8896252&partnerID=40&md5=0687e1552ed06a36aa3b0df6ea267f00
http://eprints.utp.edu.my/24888/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Advancements in wireless body sensor technology have enabled continuous recording of Electroencephalogram (EEG) data for remote monitoring. However, a significant amount of data introduced due to the continuous data recording over time has become a challenge for energy constraint sensor nodes to transfer the data to the remote stations. Therefore, many researchers explore data compression techniques to solve the large-scale data issue by compressing before the raw data are transmitted to the sink. This paper proposes a Truncated Singular Value Decomposition (TSVD) technique to compress raw EEG data by eliminating the high volume of redundant data. At the pre-processing stage, collected EEG data are reshaped to a 2-D matrix then the matrix is transformed into the subspace or vector-space using TSVD for to compress the matrix based on the correlation of the data. Afterwards, the proposed technique reconstructs the compressed data at the remote station for further analysis. Various performance metrics are utilized to evaluate the proposed technique. Simulation results show that the proposed technique suppresses a big amount of redundant data with acceptable distortion of the original data. © 2019 IEEE.