Predicting Methodology for Driver Distraction in IVIS Using ANN

G. Abinaya, A. Venkata Sai Krishna, P. Bala Saravan Teja, Kandula Venkata Sai Harsha and M. Sai Ajay

In present day scenario, due to the advancement in technology there is a high possibility that the drivers often do multi-tasking. Recent scientific results prove that these are some of the reasons for accidents. Because of performing some secondary tasks like listening music, talking on phone etc., concentration of the driver often gets affected and leads to accidents. To ensure the safety of driver and also the vehicle a system to measure the performance of the driver was developed. In this paper, we are considering these parameters like alcohol consumption level in addition to behavioural, performance-based and psychological distractions that enhance the safety measures of the driver. This system also tries to report the alcohol level if more than permissible level so, that appropriate action can be taken on the driver. The alcohol reporting process is a triggering process. To evaluate the distraction level artificial neural networks are used. So, in order to optimise the system, it uses the ED method to measure the path detection and speed of the vehicle. First the dataset has to be recorded for an ideal driver who does normal driving. So the system will be trained based on this, then series of drivers who perform secondary tasks is conducted. When a new driver is recognised, the system evaluates the features of the new driver to the ideal dataset parameters and the performance is evaluated.

Volume 11 | 04-Special Issue

Pages: 1093-1098