Back to search

Clemson University

Basic Safety Message Generation Through a Video-Based Analytics for Potential Safety Application

Abstract

dc:description.abstract

With the advancement of modern artificial intelligence techniques, computer vision can play a vital role in enhancing roadway safety by reducing the risk of imminent collisions. To do so, a vision-based safety application is required, where a roadside camera can monitor the roadway traffic and predict potential risks of crashes in real-time. If any risky situation or behavior is observed that may lead to a crash, then a safety application can send warnings to the vehicles at risk. For vision-based safety applications on a roadway section, it is important to accurately monitor each vehicle’s location, speed, acceleration, heading direction, etc. In this study, the author develops a video analytics-based basic safety message (BSM) generation method in accordance with the Society of Automotive Engineers standards (SAE J2945 and SAE J2735). The BSM generated from the algorithms developed from this research is further evaluated by conducting a field test where the results are compared with the ground truth results and cellular vehicle-to-everything (C-V2X) communication device-generated results. Our results demonstrate that our proposed video-based BSM generation method outperforms the C-V2X generated results, and our method’s errors are less than the maximum acceptable errors set by SAE J2945. Additionally, the author conducts tests to assess the end-to-end latency of our developed method and found that the end-to-end latency is within the maximum allowable range for potential safety applications. The author further proposes use case scenarios, illustrating how our BSM generation method developed through this research can be utilized for potential safety applications.

Degree

thesis:*
Name thesis:degree_name
Master of Science (MS)
Level thesis:degree_level
Thesis
Discipline thesis:degree_discipline
Civil Engineering

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Enan, Abyad
Contributors dc:contributor
  • Dr. Mashrur Chowdhury
  • Dr. M. Z. Naser
  • Dr. Siyu Huang

Subjects

dc:subject × 10

Identifiers

dc:identifier.*
Repository record dc:identifier
https://open.clemson.edu/all_theses/4390

Chain of custody

source
Harvested from
Clemson University
Base URL
open.clemson.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Enan, Abyad. Basic Safety Message Generation Through a Video-Based Analytics for Potential Safety Application. Thesis thesis, https://open.clemson.edu/all_theses/4390