"Pedestrian Early Collision Detection System"
 
Pedestrian Early Collision Detection System

Pedestrian Early Collision Detection System

Presenter(s)

Jayanth Merakanapalli, Jayanth Paturi, Sushanth Singireddy

Comments

11:00-11:20, LTC Studio

Description

Pedestrian safety is a growing concern in urban and suburban areas, where pedestrian-vehicle collisions can result in serious injuries. This research focuses on developing an early collision detection system that leverages dash cam video footage to predict potential pedestrian-vehicle accidents before they occur. The system utilizes advanced video processing techniques, including zooming, center distortion correction, and cropping, to enhance detection accuracy. By analyzing frame-by-frame motion and trajectory patterns, the system assesses the risk of a collision and provides timely alerts to drivers, allowing them to take preventive action.A key feature of the proposed system is pedestrian trajectory prediction, which estimates a pedestrian’s future movement based on historical motion data. The model evaluates pedestrian paths relative to approaching vehicles and determines the likelihood of an imminent collision. The study is conducted using a dataset of dash cam videos capturing real-world pedestrian-road interactions, where vehicles stop before a collision occurs.The research aims to contribute to the advancement of driver-assistance technologies by enhancing collision prevention mechanisms. By integrating pedestrian trajectory prediction with dash cam-based monitoring, the system provides an additional layer of safety for drivers and pedestrians.

Publication Date

4-23-2025

Project Designation

Graduate Research

Primary Advisor

Tam Nguyen

Primary Advisor's Department

Computer Science

Keywords

Stander Symposium, College of Arts and Sciences

Institutional Learning Goals

Practical Wisdom; Critical Evaluation of Our Times; Community

Pedestrian Early Collision Detection System

Share

COinS