Zapisane w:
| 1. autor: | |
|---|---|
| Format: | Recurso digital |
| Język: | |
| Wydane: |
Zenodo
2025
|
| Dostęp online: | https://doi.org/10.5281/zenodo.17577254 |
| Etykiety: |
Dodaj etykietę
Nie ma etykietki, Dołącz pierwszą etykiete!
|
Spis treści:
- <p>Urban traffic management is a major challenge due to the continuous increase in the number of vehicles, <br>leading to congestion, delays, greenhouse gas emissions, and a decline in quality of life. Traditional traffic lights, based on <br>fixed timers or inductive loops, are no longer suitable for efficiently managing complex and changing urban traffic. <br>Artificial intelligence (AI), particularly convolutional neural networks (CNNs), offers a promising solution. These <br>technologies enable real-time traffic data analysis, learning complex patterns, and adapting traffic light control to the <br>specific needs of each intersection, thus ensuring dynamic and responsive traffic management. <br>This article presents an innovative approach using a CNN model to detect traffic jams in real-time from road surveillance <br>camera data, enabling intelligent decision-making for traffic light regulation. First, we developed a CNN model capable of <br>detecting traffic jams in images. Then, we implemented the proposed model and discussed the results obtained. The model <br>demonstrated a performance of 99% on training data and 97% on test data. Its deployment in real-world conditions also <br>yielded conclusive results.</p>