Traffic Congestion Estimation using Video without Vehicle Tracking

Authors

  • Antony A. Siswoyo Institut Teknologi Bandung, Indonesia Author
  • Endra Joelianto Institut Teknologi Bandung, Indonesia Author
  • Herman Y. Sutarto PT. Pusat Riset Energi, Indonesia Author

Keywords:

Traffic control, Traffic sensor, Image thresholding, Expectation maximization, Congestion estimation, Vehicle tracking

Abstract

In this study, a traffic density monitoring system using the expectation maximization (EM) algorithm was tested on video data with varying traffic density levels. Experiments were conducted to find the most accurate image thresholding method to preprocess the images before they are fed to the EM algorithm. The algorithm successfully detected traffic density, separating it into two categories, namely ‘congested’ and ‘smooth’. Using the Bradley-Roth method for image thresholding produced the most accurate results.

Downloads

Published

2022-06-01

How to Cite

Traffic Congestion Estimation using Video without Vehicle Tracking. (2022). Internetworking Indonesia Journal, 14(1), 33-37. https://internetworkingindonesia.org/index.php/iij/article/view/30