GUHA Method Analysis of Burg's and Welch's EEG Signals Spectrograms

Abstract

The paper presents results of GUHA method analyses of electroencephalographic spectrograms (EEG spectrograms). The goal of analyses is to detect those characteristics in EEG spectrograms which are significantly related to hazardous episodes in car driver's attention. Namely, we are interested in the micro-sleeps detection on the basis of scan and analysis of an EEG signal measured on driver's head. The presented research postpones and refines our previous research in this area due to the introduction of new, more extensive experimental data set of EEG records; and employment of Burg's and Welch's transformations from time to frequency domain. Obtained results are promising since we are able to identified factors that discriminate hazardous states from non-hazardous ones.

Categories and tags

  • Kapitola v monografii