Improving driving ability using biofeedback by monitoring the mental situation by RGB camera

Abstract

In this study, we propose to improve driving ability by biofeedback using an RGB camera by monitoring the mental situation of driver. Driving a car or other vehicle while distracted is extremely dangerous. For example, distracted driving may lead to an accident because of the delay in noticing pedestrians. We believe that it is important for drivers to understand and improve their own distractibility in order to prevent accidents. However, it is difficult to estimate whether a driver is distracted or not, and to improve the situation. Therefore, it is expected to build a method for drivers to understand and improve their own attentiveness. Distractibility is known to be related to sympathetic arousal. In this study, we obtained pulse waves from facial video images using an RGB camera and estimated sympathetic nervous system values from the pulse waves. Biofeedback is provided by optical signals in real time. Biofeedback is performed by presenting color using threshold values based on the estimated sympathetic nerve values. To validate the proposed method, we generated a distracted state by subjecting him to a task and compared the results with and without biofeedback. As a result, it is shown that the amount of time spent in the distracted state with biofeedback was reduced compared to that without biofeedback.