AUTOMATIC DETECTION OF DRIVER IMPAIRMENT BASED ON PUPILLARY LIGHT REFLEX

Abstract

The main objective of this paper is to determine the feasibility of designing a driver drunkenness detection system based on the dynamic analysis of a subject’s pupillary light reflex (PLR). This involuntary reaction is widely utilized in the medical field to diagnose a variety of diseases, and in this paper, the effectiveness of such a method to reveal an impairment condition due to alcohol abuse is evaluated. The test method consists in applying a light stimulus to one eye of the subject and to capture the dynamics of constriction of both eyes; for extracting the pupil size profiles from the video sequences, a two-step methodology is described, where in the first phase, the iris/pupil search within the image is performed, and in the second stage, the image is cropped to perform pupil detection on a smaller image to improve time efficiency. The undesired pupil dynamics arising in the PLR are defined and evaluated; a spontaneous oscillation of the pupil diameter is observed in the range [0, 2] Hz and the accommodation reflex causes pupil constriction of about 10% of the iris diameter. A database of pupillary light responses is acquired on different subjects in baseline condition and after alcohol consumption, and for each one, a first-order model is identified. A set of features is introduced to compare the two populations of responses and is used to design a support vector machine classifier to discriminate between “Sober” and “Drunk” states.

Existing System

Convolutional Neural Networks (CNN), Multi-Task Cascaded Convolutional Network (MTCNN)

Proposed System

The contribution of this paper is threefold. First, a time efficient methodology is proposed for processing video sequences and extract a time profile of the pupil diameter, consisting in two steps. In the first one, the region of the image containing the eye is detected and, in parallel, iris size is measured until detection stabilization; in the second step, the image is cropped to contain only the pupil, which allows for a more time-efficient pupil size measurement. The cropped region position is constantly adapted in order to compensate for undesired eye/face movements. Second, an analysis on different subjects is performed to observe the pupil size variation in absence of external stimulation; in this analysis, the typical phenomena that are involved in the pupil dynamics in absence of external stimulation are observed and quantified and some considerations on the inter subject variability of the static values of the pupil size are made. Third, the paper assesses the problem of determining whether it is feasible to detect the drunkenness state of a subject based on the dynamic characteristics of his/her pupillary light response. In order to perform such study, a light stimulus is applied to one eye of different subjects and the pupillary response of both eyes is recorded through video cameras. Experimental data are collected from different subjects in baseline condition and after alcohol assumption; for each observation, a simple model describing the pupil constriction dynamics is identified, whose parameters are considered as features. Finally, a SVM-based classifier is designed on such features to discriminate between “Sober” and “Drunk” subjects.  

Architecture

BLOCK DIAGRAM