A Low-Cost System for Dynamic Analysis of Pupillary Light Response for a Driver Drowsiness Detection System

This paper proposes a low-cost methodology for analyzing the dynamics of constriction of a human pupil while subjected to a light stimulus: this phoenomenon is commonly known as Pupillary Light Reflex (PLR) and is widely utilized in medical field to diagnose a variety of diseases. In particular, the analysis of the PLR in this paper is preparatory to the development of a Driver Drowsiness Detection System (DDDS), which reveals the driver's sleepiness state by measuring the pupil's constriction dynamics. The test protocol consists in applying a light stimulus to one eye of the subject and to capture the dynamics of constriction of both eyes through cameras; the proposed methodology extracts from the video sequences the time profile of the pupil diameter, from which dynamic and static features are obtained by fitting a simplified $1^{st}$-order model with delay. Finally, conclusions on the intra-and inter-subject variability of such features are drawn and possible DDDS strategies are proposed based on the obtained results.