KIT | KIT-Bibliothek | Impressum | Datenschutz

Automatic analysis of UAS-based thermal images to detect leakages in district heating systems

Vollmer, Elena 1; Volk, Rebekka ORCID iD icon 1; Schultmann, Frank ORCID iD icon 1
1 Institut für Industriebetriebslehre und Industrielle Produktion (IIP), Karlsruher Institut für Technologie (KIT)

Abstract (englisch):

The mostly subterranean nature of district heating system pipelines makes pinpointing any occurring leakages a challenge. Airborne thermography offers a means for widespread monitoring, allowing thermal anomalies to be identified within multitudes of infrared images. This paper details a program developed to automate the entire image analysis process using mainly open-source software. Thermal images are acquired via unmanned aircraft system (UAS), pre-processed, and georeferenced individually or combined to orthomosaics. The search space is minimized to areas around the pipelines. Regions of interest are determined by image segmentation via tailored triangle histogram thresholding. The majority of resulting false alarms are removed by comparison with characteristic traits and results classified by their severity. The algorithm is applied to images newly acquired in Germany as part of a case study. The implemented methodology allows for a reduction of between 92 and 99% of thermal anomalies to a manageable amount of potential leakages for network operators to view. The use of orthomosaiking software in this context, though helpful in coalescing data, is found to lack robustness, precision and therefore reliability. ... mehr


Postprint §
DOI: 10.5445/IR/1000163771
Veröffentlicht am 26.10.2024
Originalveröffentlichung
DOI: 10.1080/01431161.2023.2242586
Scopus
Zitationen: 3
Web of Science
Zitationen: 1
Dimensions
Zitationen: 3
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Industriebetriebslehre und Industrielle Produktion (IIP)
Publikationstyp Zeitschriftenaufsatz
Publikationsjahr 2023
Sprache Englisch
Identifikator ISSN: 0143-1161, 1366-5901
KITopen-ID: 1000163771
Erschienen in International Journal of Remote Sensing
Verlag Taylor and Francis
Seiten 31 S.
Vorab online veröffentlicht am 25.10.2023
Schlagwörter District heating systems; image processing; leakage detection; remote thermography; software development; unmanned aircraft systems
Nachgewiesen in Dimensions
Web of Science
Scopus
Globale Ziele für nachhaltige Entwicklung Ziel 7 – Bezahlbare und saubere EnergieZiel 11 – Nachhaltige Städte und Gemeinden
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page