Edge-guided slicing and inference (EGSI) with feature aggregation network for quantification of large defective temporary construction materials
Temporary materials, primarily steel tubular sections, are essential for erecting temporary structures on construction sites. Due to frequent reuse and storage conditions, these materials often develop defects such as rust and bends, compromising their safety. While manual quantification of defectiv...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2025-07-01
|
| Series: | Journal of Asian Architecture and Building Engineering |
| Subjects: | |
| Online Access: | http://dx.doi.org/10.1080/13467581.2025.2523585 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849319122546982912 |
|---|---|
| author | Njoroge James Mugo Akbar Ali Song Jinwoo Soonwook Kwon |
| author_facet | Njoroge James Mugo Akbar Ali Song Jinwoo Soonwook Kwon |
| author_sort | Njoroge James Mugo |
| collection | DOAJ |
| description | Temporary materials, primarily steel tubular sections, are essential for erecting temporary structures on construction sites. Due to frequent reuse and storage conditions, these materials often develop defects such as rust and bends, compromising their safety. While manual quantification of defective materials is time-consuming, especially for batches exceeding 200 materials, computer vision techniques offer a more efficient alternative. However, detecting defects in large batches remains challenging as the materials appear as small objects in images. This paper proposes a feature aggregation network through an ablation study to enhance the detection accuracy of large defective temporary materials and an edge-guided slicing inference method for precise quantification of large batch sizes. The proposed bottleneck layer in the feature aggregation network achieved a mean average precision (mAP) of 81.1%, outperforming the best custom model by 1.2%. Additionally, the edge-guided slicing inference reduced the mean absolute error to 0.1%, compared to 0.6% from the original slicing-aided hyper inference. The developed system was also deployed on a web-based user interface for visualization, improving accessibility and usability. These advancements enhance automated defect detection and quantification of temporary materials, improving efficiency and accuracy in construction management. |
| format | Article |
| id | doaj-art-be9a763c2cc04264ab4489ce4a1968e8 |
| institution | Kabale University |
| issn | 1347-2852 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Journal of Asian Architecture and Building Engineering |
| spelling | doaj-art-be9a763c2cc04264ab4489ce4a1968e82025-08-20T03:50:38ZengTaylor & Francis GroupJournal of Asian Architecture and Building Engineering1347-28522025-07-010011710.1080/13467581.2025.25235852523585Edge-guided slicing and inference (EGSI) with feature aggregation network for quantification of large defective temporary construction materialsNjoroge James Mugo0Akbar Ali1Song Jinwoo2Soonwook Kwon3Sungkyunkwan UniversitySungkyunkwan UniversitySeoul National UniversitySungkyunkwan UniversityTemporary materials, primarily steel tubular sections, are essential for erecting temporary structures on construction sites. Due to frequent reuse and storage conditions, these materials often develop defects such as rust and bends, compromising their safety. While manual quantification of defective materials is time-consuming, especially for batches exceeding 200 materials, computer vision techniques offer a more efficient alternative. However, detecting defects in large batches remains challenging as the materials appear as small objects in images. This paper proposes a feature aggregation network through an ablation study to enhance the detection accuracy of large defective temporary materials and an edge-guided slicing inference method for precise quantification of large batch sizes. The proposed bottleneck layer in the feature aggregation network achieved a mean average precision (mAP) of 81.1%, outperforming the best custom model by 1.2%. Additionally, the edge-guided slicing inference reduced the mean absolute error to 0.1%, compared to 0.6% from the original slicing-aided hyper inference. The developed system was also deployed on a web-based user interface for visualization, improving accessibility and usability. These advancements enhance automated defect detection and quantification of temporary materials, improving efficiency and accuracy in construction management.http://dx.doi.org/10.1080/13467581.2025.2523585convolutional neural networkconstruction materialsconstruction site inspectiondefect detection |
| spellingShingle | Njoroge James Mugo Akbar Ali Song Jinwoo Soonwook Kwon Edge-guided slicing and inference (EGSI) with feature aggregation network for quantification of large defective temporary construction materials Journal of Asian Architecture and Building Engineering convolutional neural network construction materials construction site inspection defect detection |
| title | Edge-guided slicing and inference (EGSI) with feature aggregation network for quantification of large defective temporary construction materials |
| title_full | Edge-guided slicing and inference (EGSI) with feature aggregation network for quantification of large defective temporary construction materials |
| title_fullStr | Edge-guided slicing and inference (EGSI) with feature aggregation network for quantification of large defective temporary construction materials |
| title_full_unstemmed | Edge-guided slicing and inference (EGSI) with feature aggregation network for quantification of large defective temporary construction materials |
| title_short | Edge-guided slicing and inference (EGSI) with feature aggregation network for quantification of large defective temporary construction materials |
| title_sort | edge guided slicing and inference egsi with feature aggregation network for quantification of large defective temporary construction materials |
| topic | convolutional neural network construction materials construction site inspection defect detection |
| url | http://dx.doi.org/10.1080/13467581.2025.2523585 |
| work_keys_str_mv | AT njorogejamesmugo edgeguidedslicingandinferenceegsiwithfeatureaggregationnetworkforquantificationoflargedefectivetemporaryconstructionmaterials AT akbarali edgeguidedslicingandinferenceegsiwithfeatureaggregationnetworkforquantificationoflargedefectivetemporaryconstructionmaterials AT songjinwoo edgeguidedslicingandinferenceegsiwithfeatureaggregationnetworkforquantificationoflargedefectivetemporaryconstructionmaterials AT soonwookkwon edgeguidedslicingandinferenceegsiwithfeatureaggregationnetworkforquantificationoflargedefectivetemporaryconstructionmaterials |