Mapping stains on flat roofs using semantic segmentation based on deep learning
Moisture stains indicate ongoing degradation processes and may reveal areas of the roof slab where water infiltration occurs, compromising the performance and durability of the building system. During inspections of roofing systems, an inspector's field of vision differs from that of drones dur...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-07-01
|
| Series: | Case Studies in Construction Materials |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2214509524012580 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846109761757511680 |
|---|---|
| author | Lara Monalisa Alves dos Santos Leonardo Rabero Lescano Gabriel Toshio Hirokawa Higa Vanda Alice Garcia Zanoni Lenildo Santos da Silva Cesar Ivan Alvarez Hemerson Pistori |
| author_facet | Lara Monalisa Alves dos Santos Leonardo Rabero Lescano Gabriel Toshio Hirokawa Higa Vanda Alice Garcia Zanoni Lenildo Santos da Silva Cesar Ivan Alvarez Hemerson Pistori |
| author_sort | Lara Monalisa Alves dos Santos |
| collection | DOAJ |
| description | Moisture stains indicate ongoing degradation processes and may reveal areas of the roof slab where water infiltration occurs, compromising the performance and durability of the building system. During inspections of roofing systems, an inspector's field of vision differs from that of drones during overflights. As a result, traditional inspections might not always detect the presence and severity of stains, making maintenance on flat roofs a complex task. In this context, this experimental study aims to analyze deep learning-based semantic segmentation with images obtained from drones to map and monitor damp patches during automated building inspections of flat roof systems. The research tested two convolutional neural networks for semantic segmentation: the Fully Convolutional Network (FCN) with a ResNet50 backbone and DeepLabV3 with a ResNet101 backbone, as well as a transformer-based deep artificial neural network called SegFormer with a MiT-B1 backbone. We evaluated three optimizers for each model—Adam, Adagrad, and SGD—along with learning rates of 1e-2, 1e-3, and 1e-4. The models were compared using four performance metrics. The FCN, optimized with Adagrad at a learning rate of 1e-2, showed the best results. The average metrics obtained in this case were as follows: precision: 79.69 %, recall: 67.81 %, F-score: 73.09 %, and Intersection over Union (IoU): 57.70 %. |
| format | Article |
| id | doaj-art-09715e0b468d4a3399bfc988e084f0b9 |
| institution | Kabale University |
| issn | 2214-5095 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Case Studies in Construction Materials |
| spelling | doaj-art-09715e0b468d4a3399bfc988e084f0b92024-12-25T04:21:19ZengElsevierCase Studies in Construction Materials2214-50952025-07-0122e04106Mapping stains on flat roofs using semantic segmentation based on deep learningLara Monalisa Alves dos Santos0Leonardo Rabero Lescano1Gabriel Toshio Hirokawa Higa2Vanda Alice Garcia Zanoni3Lenildo Santos da Silva4Cesar Ivan Alvarez5Hemerson Pistori6University of Brasilia, Distrito Federal, Brazil; Corresponding author.Dom Bosco Catholic University, Mato Grosso do Sul, BrazilDom Bosco Catholic University, Mato Grosso do Sul, BrazilUniversity of Brasilia, Distrito Federal, BrazilUniversity of Brasilia, Distrito Federal, BrazilCentre for Climate Resilience, University of Augsburg, Universitätsstrasse 12a, Augsburg 86159, Germany; Environmental Research Group for Sustainable Development (GIADES), Salesian Polytechnic University, Rumichaca y Moran Valverde, Quito, EcuadorDom Bosco Catholic University, Mato Grosso do Sul, Brazil; Federal University of Mato Grosso do Sul, Mato Grosso do Sul, BrazilMoisture stains indicate ongoing degradation processes and may reveal areas of the roof slab where water infiltration occurs, compromising the performance and durability of the building system. During inspections of roofing systems, an inspector's field of vision differs from that of drones during overflights. As a result, traditional inspections might not always detect the presence and severity of stains, making maintenance on flat roofs a complex task. In this context, this experimental study aims to analyze deep learning-based semantic segmentation with images obtained from drones to map and monitor damp patches during automated building inspections of flat roof systems. The research tested two convolutional neural networks for semantic segmentation: the Fully Convolutional Network (FCN) with a ResNet50 backbone and DeepLabV3 with a ResNet101 backbone, as well as a transformer-based deep artificial neural network called SegFormer with a MiT-B1 backbone. We evaluated three optimizers for each model—Adam, Adagrad, and SGD—along with learning rates of 1e-2, 1e-3, and 1e-4. The models were compared using four performance metrics. The FCN, optimized with Adagrad at a learning rate of 1e-2, showed the best results. The average metrics obtained in this case were as follows: precision: 79.69 %, recall: 67.81 %, F-score: 73.09 %, and Intersection over Union (IoU): 57.70 %.http://www.sciencedirect.com/science/article/pii/S2214509524012580MaintenanceBuilding inspectionComputer VisionCNN – FCN and DeepLabV3Transformer – SegFormer |
| spellingShingle | Lara Monalisa Alves dos Santos Leonardo Rabero Lescano Gabriel Toshio Hirokawa Higa Vanda Alice Garcia Zanoni Lenildo Santos da Silva Cesar Ivan Alvarez Hemerson Pistori Mapping stains on flat roofs using semantic segmentation based on deep learning Case Studies in Construction Materials Maintenance Building inspection Computer Vision CNN – FCN and DeepLabV3 Transformer – SegFormer |
| title | Mapping stains on flat roofs using semantic segmentation based on deep learning |
| title_full | Mapping stains on flat roofs using semantic segmentation based on deep learning |
| title_fullStr | Mapping stains on flat roofs using semantic segmentation based on deep learning |
| title_full_unstemmed | Mapping stains on flat roofs using semantic segmentation based on deep learning |
| title_short | Mapping stains on flat roofs using semantic segmentation based on deep learning |
| title_sort | mapping stains on flat roofs using semantic segmentation based on deep learning |
| topic | Maintenance Building inspection Computer Vision CNN – FCN and DeepLabV3 Transformer – SegFormer |
| url | http://www.sciencedirect.com/science/article/pii/S2214509524012580 |
| work_keys_str_mv | AT laramonalisaalvesdossantos mappingstainsonflatroofsusingsemanticsegmentationbasedondeeplearning AT leonardoraberolescano mappingstainsonflatroofsusingsemanticsegmentationbasedondeeplearning AT gabrieltoshiohirokawahiga mappingstainsonflatroofsusingsemanticsegmentationbasedondeeplearning AT vandaalicegarciazanoni mappingstainsonflatroofsusingsemanticsegmentationbasedondeeplearning AT lenildosantosdasilva mappingstainsonflatroofsusingsemanticsegmentationbasedondeeplearning AT cesarivanalvarez mappingstainsonflatroofsusingsemanticsegmentationbasedondeeplearning AT hemersonpistori mappingstainsonflatroofsusingsemanticsegmentationbasedondeeplearning |