Möjligheter och begränsningar med termografi som metod för broinspektion
(2025) In CODEN: LUTVDG/(TVTT-5373)/1-77/2025 VGMM01 20242Transport and Roads
- Abstract
- This study examines the opportunities and limitations of thermography as a method for bridge inspections, focusing on its ability to identify defects and moisture in concrete structures. The method is evaluated based on the types of defects that can be detected and how environmental factors influence the results. Furthermore, the study explores the potential of using machine learning to automatically detect defects in thermal images.
The methodology includes data collection through field inspections performed with a handheld thermal camera. Measurements were taken at different times and under varying weather conditions to assess the method’s sensitivity to factors such as temperature, humidity, and solar radiation. Additionally, a... (More) - This study examines the opportunities and limitations of thermography as a method for bridge inspections, focusing on its ability to identify defects and moisture in concrete structures. The method is evaluated based on the types of defects that can be detected and how environmental factors influence the results. Furthermore, the study explores the potential of using machine learning to automatically detect defects in thermal images.
The methodology includes data collection through field inspections performed with a handheld thermal camera. Measurements were taken at different times and under varying weather conditions to assess the method’s sensitivity to factors such as temperature, humidity, and solar radiation. Additionally, a machine learning-based image classification model and a segmentation model were trained on the collected data to evaluate their ability to automatically identify damaged areas.
The results indicate that thermography can be a useful complement to bridge inspections, although its accuracy is influenced by external factors. The measurements demonstrated that the thermal camera is capable of detecting visible cracks, hidden defects, and moisture accumulations, but also revealed limitations, as some defects were not always detected. Solar radiation and humidity were identified as the most significant environmental factors and should therefore be considered during measurements to ensure reliable outcomes. The results also show that machine learning has the potential to automate detection of defects, and with additional resources, more efficient and reliable models could be developed.
Lastly, recommendations are presented on how thermography should be applied in bridge inspections to ensure reliable and effective measurements. (Less) - Abstract (Swedish)
- Denna studie undersöker möjligheter och begränsningar med termografi som metod för broinspektion, med fokus på dess förmåga att identifiera skador och fukt i betongkonstruktioner. Metoden utvärderas både utifrån vilka typer av skador som kan detekteras och hur omgivningsfaktorer påverkar mätresultaten. Vidare undersöks potentialen i att använda maskininlärning för att automatiskt identifiera skador i termiska bilder.
Metoden omfattar datainsamling i form av fältundersökningar, utförda med en handhållen värmekamera. Mätningarna har skett under varierande tider och väderförhållanden för att studera metodens känslighet för faktorer som temperatur, luftfuktighet och solstrålning. Dessutom har en bildklassificeringsmodell och en... (More) - Denna studie undersöker möjligheter och begränsningar med termografi som metod för broinspektion, med fokus på dess förmåga att identifiera skador och fukt i betongkonstruktioner. Metoden utvärderas både utifrån vilka typer av skador som kan detekteras och hur omgivningsfaktorer påverkar mätresultaten. Vidare undersöks potentialen i att använda maskininlärning för att automatiskt identifiera skador i termiska bilder.
Metoden omfattar datainsamling i form av fältundersökningar, utförda med en handhållen värmekamera. Mätningarna har skett under varierande tider och väderförhållanden för att studera metodens känslighet för faktorer som temperatur, luftfuktighet och solstrålning. Dessutom har en bildklassificeringsmodell och en bildsegmenteringsmodell tränats på insamlad data för att undersöka deras förmåga att identifiera skadeområden automatiskt.
Resultatet visar att termografi kan vara ett användbart komplement vid broinspektioner, men att dess noggrannhet påverkas av omgivningsfaktorer. Mätningarna visade att värmekameran kan detektera synliga sprickor, dolda skador och fuktsamlingar, men också att den har begränsningar, då vissa skador inte alltid kunde detekteras. Solinstrålning och luftfuktighet identifierades som de mest betydande omgivningsfaktorerna och bör beaktas vid mätningar för att säkerställa tillförlitliga resultat. Resultatet visar även att maskininlärning har potential att automatisera identifieringen av skador, och med mer resurser kan effektiva och mer tillförlitliga modeller utvecklas.
Avslutningsvis presenteras rekommendationer för hur termografi bör tillämpas vid broinspektioner för att säkerställa pålitliga och effektiva mätningar. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9188736
- author
- Ahlqvist, Emil LU
- supervisor
- organization
- alternative title
- Capabilities and limitations of thermography as a method for bridge inspection
- course
- VGMM01 20242
- year
- 2025
- type
- H3 - Professional qualifications (4 Years - )
- subject
- keywords
- Termografi, Broinspektion, Värmekamera, Bildklassificering, Bildsegmentering
- publication/series
- CODEN: LUTVDG/(TVTT-5373)/1-77/2025
- report number
- 406
- ISSN
- 1653-1922
- language
- Swedish
- id
- 9188736
- date added to LUP
- 2025-05-27 11:11:39
- date last changed
- 2025-05-27 11:11:39
@misc{9188736,
abstract = {{This study examines the opportunities and limitations of thermography as a method for bridge inspections, focusing on its ability to identify defects and moisture in concrete structures. The method is evaluated based on the types of defects that can be detected and how environmental factors influence the results. Furthermore, the study explores the potential of using machine learning to automatically detect defects in thermal images.
The methodology includes data collection through field inspections performed with a handheld thermal camera. Measurements were taken at different times and under varying weather conditions to assess the method’s sensitivity to factors such as temperature, humidity, and solar radiation. Additionally, a machine learning-based image classification model and a segmentation model were trained on the collected data to evaluate their ability to automatically identify damaged areas.
The results indicate that thermography can be a useful complement to bridge inspections, although its accuracy is influenced by external factors. The measurements demonstrated that the thermal camera is capable of detecting visible cracks, hidden defects, and moisture accumulations, but also revealed limitations, as some defects were not always detected. Solar radiation and humidity were identified as the most significant environmental factors and should therefore be considered during measurements to ensure reliable outcomes. The results also show that machine learning has the potential to automate detection of defects, and with additional resources, more efficient and reliable models could be developed.
Lastly, recommendations are presented on how thermography should be applied in bridge inspections to ensure reliable and effective measurements.}},
author = {{Ahlqvist, Emil}},
issn = {{1653-1922}},
language = {{swe}},
note = {{Student Paper}},
series = {{CODEN: LUTVDG/(TVTT-5373)/1-77/2025}},
title = {{Möjligheter och begränsningar med termografi som metod för broinspektion}},
year = {{2025}},
}