Application of YOLO and U-Net models for building material identification on segmented images
This paper is devoted to the analysis of existing convolutional neural networks and experimental verification of the YOLO and U-Net architectures for the identification and classification of building materials based on images of destroyed structures. The aim of the study is to determine the effecti...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Lublin University of Technology
2025-06-01
|
| Series: | Informatyka, Automatyka, Pomiary w Gospodarce i Ochronie Środowiska |
| Subjects: | |
| Online Access: | https://ph.pollub.pl/index.php/iapgos/article/view/6968 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849419776217055232 |
|---|---|
| author | Ruslan Voronkov Mykhailo Bezugliy |
| author_facet | Ruslan Voronkov Mykhailo Bezugliy |
| author_sort | Ruslan Voronkov |
| collection | DOAJ |
| description |
This paper is devoted to the analysis of existing convolutional neural networks and experimental verification of the YOLO and U-Net architectures for the identification and classification of building materials based on images of destroyed structures. The aim of the study is to determine the effectiveness of these models in the tasks of recognising materials suitable for reuse and recycling. This will help reduce construction waste and introduce a more environmentally friendly approach to resource management. The study examined several modern deep learning models for image processing, including Faster R-CNN, Mask R-CNN, FCN (Fully Convolutional Networks), and SegNet. However, the choice was made on the YOLO and U-Net architectures. YOLO is used for fast object identification in images, which allows for quick detection and classification of building materials, and U-Net is used for detailed image segmentation, providing accurate determination of the structure and composition of building materials. Each of these models has been adapted to the specific requirements of building materials analysis in the context of collapsed structures. Experimental results have shown that the use of these models allows achieving high accuracy of segmentation of images of destroyed buildings, which makes them promising for use in automated resource control systems.
|
| format | Article |
| id | doaj-art-8cf836759cda4b3e9b35dc051f32a101 |
| institution | Kabale University |
| issn | 2083-0157 2391-6761 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Lublin University of Technology |
| record_format | Article |
| series | Informatyka, Automatyka, Pomiary w Gospodarce i Ochronie Środowiska |
| spelling | doaj-art-8cf836759cda4b3e9b35dc051f32a1012025-08-20T03:31:57ZengLublin University of TechnologyInformatyka, Automatyka, Pomiary w Gospodarce i Ochronie Środowiska2083-01572391-67612025-06-0115210.35784/iapgos.6968Application of YOLO and U-Net models for building material identification on segmented imagesRuslan Voronkov0https://orcid.org/0009-0000-4779-0132Mykhailo Bezugliy1https://orcid.org/0000-0003-0624-0585National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute"National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" This paper is devoted to the analysis of existing convolutional neural networks and experimental verification of the YOLO and U-Net architectures for the identification and classification of building materials based on images of destroyed structures. The aim of the study is to determine the effectiveness of these models in the tasks of recognising materials suitable for reuse and recycling. This will help reduce construction waste and introduce a more environmentally friendly approach to resource management. The study examined several modern deep learning models for image processing, including Faster R-CNN, Mask R-CNN, FCN (Fully Convolutional Networks), and SegNet. However, the choice was made on the YOLO and U-Net architectures. YOLO is used for fast object identification in images, which allows for quick detection and classification of building materials, and U-Net is used for detailed image segmentation, providing accurate determination of the structure and composition of building materials. Each of these models has been adapted to the specific requirements of building materials analysis in the context of collapsed structures. Experimental results have shown that the use of these models allows achieving high accuracy of segmentation of images of destroyed buildings, which makes them promising for use in automated resource control systems. https://ph.pollub.pl/index.php/iapgos/article/view/6968image segmentationneural networksclassification of building materialsYOLOv8U-Netdeep learning |
| spellingShingle | Ruslan Voronkov Mykhailo Bezugliy Application of YOLO and U-Net models for building material identification on segmented images Informatyka, Automatyka, Pomiary w Gospodarce i Ochronie Środowiska image segmentation neural networks classification of building materials YOLOv8 U-Net deep learning |
| title | Application of YOLO and U-Net models for building material identification on segmented images |
| title_full | Application of YOLO and U-Net models for building material identification on segmented images |
| title_fullStr | Application of YOLO and U-Net models for building material identification on segmented images |
| title_full_unstemmed | Application of YOLO and U-Net models for building material identification on segmented images |
| title_short | Application of YOLO and U-Net models for building material identification on segmented images |
| title_sort | application of yolo and u net models for building material identification on segmented images |
| topic | image segmentation neural networks classification of building materials YOLOv8 U-Net deep learning |
| url | https://ph.pollub.pl/index.php/iapgos/article/view/6968 |
| work_keys_str_mv | AT ruslanvoronkov applicationofyoloandunetmodelsforbuildingmaterialidentificationonsegmentedimages AT mykhailobezugliy applicationofyoloandunetmodelsforbuildingmaterialidentificationonsegmentedimages |