A Mobile Drowsiness Detection System with Aid of Real-Time EOG Monitoring and Infrared Ray Imaging

Although many bio-electrical methodologies were widely applied drowsiness, yet lacked of definitive criterions in investigating early-detection of fatigue state. Thus, this project combined two different non-invasive measurements, i.e. real-time electrooculography (EOG) recording and video surveillance, to qualitatively and quantitatively evaluate physiological condition during drowsy onset of normal adults. To achieve simultaneously monitoring, besides home-made EOG circuits measurement, we adopted hat-mounted CCD with infrared ray (IR) source to avoid eye-tracking and non-uniformly lighting problems. Again, simplified image processing algorithm was also crucial for the performance of embedded FPGA framework. Our preliminary results showed that this monitoring system worked well in first about 30 minutes, however lagging problem of image processing became serious as time elapsed