NACH OBEN

Firdes Celik

M. Sc.
Wissenschaftliche Mitarbeiterin
Raum: IC 6/69
Tel.: 0234 32 - 28983
E-Mail

Digitalization of Bridge Inspections and Bridge Damage Documentation with Machine Learning

Engineers regularly inspect bridges for damage to ensure their structural stability and traffic safety. However, the manual process for documenting damage is often error-prone and inconsistent, particularly in terms of measurement, description, and evaluation. Therefore, my research focuses on developing an image-based machine learning pipeline that automates on-site damage documentation. This pipeline includes damage recognition, description, and evaluation, with the goal of improving accuracy, consistency, and efficiency.

Bachelor

  • Duski, Husan (2021): Automatische Erkennung von Kiesnestern und Abplatzungen mithilfe eines maschinellen Klassifizierungsverfahrens
  • Kozlova, Anastasiia (2022): A Comparative Study of Transfer Learning Methods for Crack Segmentation
  • Baumann, Michael William (2023): Teilautomatisierte Dokumentation von Feuchtigkeitsschäden an Brücken mithilfe des maschinellen Lernens

Master

  • Kunz, Artur (2021): Deep-Learning-basierte 3D-Risserkennung in Punktwolken
  • Jiang, Sing Da (2022): Solving Data Imbalance Problems in Road Damage Detection
  • Matzke, Stephan (2022): Optimization of a Pallet Block Re-Identification System Based on Data Augmentation with Generative Adversarial Networks

Publikationen