Original title:
DBSCAN Application for Bubble Tracking in Aerated Columns
Authors:
Shirokov, Vladislav Document type: Papers Conference/Event: Bažant Postgraduate Conference 2025, Prague (CZ), 20250617
Year:
2025
Language:
eng Abstract:
In this work, DBSCAN is used not for general clustering, but specifically to reconstruct individual bubble trajectories by grouping spatially\nand temporally related position data. Each trajectory is formed as a separate cluster based on the density of detected bubble positions in space (2D) and time. The algorithm does not require prior knowledge of the number of trajectories and can adapt to the irregular motion typical for bubbles in gas–liquid systems. Host item entry: Proceedings of Abstracts, ISBN 978-80-86186-35-1
Rights: This work is protected under the Copyright Act No. 121/2000 Coll.