Automatic Classification of Sober/Drunk through Iris Pattern Recognition

In this paper, an automatic system is proposed for classification of sober or drunk by analyzing the Iris patterns of eyes. This mainly aims at developing a system which can prevent drunk driving in order to curb the road accidents caused due to drunk driving. Before starting the vehicles, drivers are made to go through iris scanning, only if they are identified as sober, then only allowed to begin the journey. A standard dataset consisting of 1275 images are used in developing the system. The proposed system is implemented by using Convolutional Neural Network (CNN) model where, the decisive characteristic of the system will be handled by using features such as the dilation of the pupil and the darkening of sclera that occurs when the person is drunk. The face of the driver is to be detected first using the Open CV modules which also localizes the eye portion in the face which is the area of interest. A better accuracy of 89% is achieved with the proposed system.