Feature Analysis of Eyeblink Waveform for Automatically Classifying Conscious Blinks

In this paper, we propose and evaluate a new feature parameter of eyeblink for automatically classifying between conscious (voluntary) and unconscious (spontaneous) blinks. We focused on integral values of an eyeblink waveform, which is defined as a measurement record of a progression of eyeblinks, as the feature parameter. Previous researchers have used duration time and waveform amplitude as the feature parameter, whereas the integral values have both feature parameter characteristics. We obtained these parameters using an NTSC video camera by splitting a single interlaced image into two fields. We used flame-splitting methods to obtain and analyze the integral value of the eyeblink waveform, and we experimentally compared the feature parameters to automatically classify the conscious and unconscious eyeblink. Duration and amplitude did not show a significant difference in some subject cases, but we confirmed a significant difference when using the integral value. Our results suggest that integral values of eyeblink waveform are effective for classifying the two eyeblink types.