DUI (Driver Under Influence) Detection Using Open CV

Abstract

Through the analysis of eye movement patterns, particularly saccadic and fixation behaviors, this study investigates a novel method for identifying driving under the influence (DUI). Intoxication has a major impact on fixations, which occur when the gaze rests on a point, and saccadic movements, which are quick eye movements between points. This approach seeks to offer a precise, non-invasive way to identify DUI by utilizing these markers. The study analyzes gaze data using sophisticated machine learning techniques to find variations in saccadic velocity, amplitude, and fixation time that are associated with drunkenness. This research has the potential to improve roadside testing techniques, decrease subjective evaluations, and increase traffic safety.