Efficient Dataset Collection for Concrete Crack Detection With Spatial-Adaptive Data Augmentation
In this paper, we propose a data augmentation technique based on Convolutional Neural Networks (CNN or ConvNet) training to efficiently obtain a dataset of images containing concrete cracks. Concrete cracks usually do not have a standardized shape and have complex patterns, making it difficult to ob...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10298197/ |
| _version_ | 1797634003110461440 |
|---|---|
| author | Jong-Hyun Kim Jung Lee |
| author_facet | Jong-Hyun Kim Jung Lee |
| author_sort | Jong-Hyun Kim |
| collection | DOAJ |
| description | In this paper, we propose a data augmentation technique based on Convolutional Neural Networks (CNN or ConvNet) training to efficiently obtain a dataset of images containing concrete cracks. Concrete cracks usually do not have a standardized shape and have complex patterns, making it difficult to obtain images of them, and there is a risk of exposure to dangerous situations when securing data. Therefore, in this paper, we efficiently address the difficulty of dataset collection by using a data augmentation technique based on learning the direction and thickness of cracks, which is cost-effective and time-efficient. Moreover, to improve efficiency, we introduce a method of adaptively handling crack data by constructing a quadtree based on the presence of cracks. To confirm the extent of the improvement in accuracy, we conducted experiments applying the crack detection algorithm to various scenes, and the accuracy was improved in all scenes when measured by IoU (Intersection over union) accuracy. When the algorithm was performed without augmenting the crack data, the false detection rate was about 25%. However, when we augmented the data using our method, the false detection rate significantly decreased to 3%. |
| first_indexed | 2024-03-11T12:02:40Z |
| format | Article |
| id | doaj.art-6c7e985e78d34a48a5b8cfd24b2ed71e |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-11T12:02:40Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-6c7e985e78d34a48a5b8cfd24b2ed71e2023-11-08T00:01:24ZengIEEEIEEE Access2169-35362023-01-011112190212191310.1109/ACCESS.2023.332824310298197Efficient Dataset Collection for Concrete Crack Detection With Spatial-Adaptive Data AugmentationJong-Hyun Kim0https://orcid.org/0000-0003-1603-2675Jung Lee1Department of Design Technology, College of Software and Convergence, Inha University, Michuhol-gu, Incheon, South KoreaDepartment of Computer Engineering, Hanbat National University, Yuseong-gu, Daejeon, South KoreaIn this paper, we propose a data augmentation technique based on Convolutional Neural Networks (CNN or ConvNet) training to efficiently obtain a dataset of images containing concrete cracks. Concrete cracks usually do not have a standardized shape and have complex patterns, making it difficult to obtain images of them, and there is a risk of exposure to dangerous situations when securing data. Therefore, in this paper, we efficiently address the difficulty of dataset collection by using a data augmentation technique based on learning the direction and thickness of cracks, which is cost-effective and time-efficient. Moreover, to improve efficiency, we introduce a method of adaptively handling crack data by constructing a quadtree based on the presence of cracks. To confirm the extent of the improvement in accuracy, we conducted experiments applying the crack detection algorithm to various scenes, and the accuracy was improved in all scenes when measured by IoU (Intersection over union) accuracy. When the algorithm was performed without augmenting the crack data, the false detection rate was about 25%. However, when we augmented the data using our method, the false detection rate significantly decreased to 3%.https://ieeexplore.ieee.org/document/10298197/Spatial-adaptive augmentationconcrete crackdata augmentationconvolutional neural networkscrack directioncrack thickness |
| spellingShingle | Jong-Hyun Kim Jung Lee Efficient Dataset Collection for Concrete Crack Detection With Spatial-Adaptive Data Augmentation IEEE Access Spatial-adaptive augmentation concrete crack data augmentation convolutional neural networks crack direction crack thickness |
| title | Efficient Dataset Collection for Concrete Crack Detection With Spatial-Adaptive Data Augmentation |
| title_full | Efficient Dataset Collection for Concrete Crack Detection With Spatial-Adaptive Data Augmentation |
| title_fullStr | Efficient Dataset Collection for Concrete Crack Detection With Spatial-Adaptive Data Augmentation |
| title_full_unstemmed | Efficient Dataset Collection for Concrete Crack Detection With Spatial-Adaptive Data Augmentation |
| title_short | Efficient Dataset Collection for Concrete Crack Detection With Spatial-Adaptive Data Augmentation |
| title_sort | efficient dataset collection for concrete crack detection with spatial adaptive data augmentation |
| topic | Spatial-adaptive augmentation concrete crack data augmentation convolutional neural networks crack direction crack thickness |
| url | https://ieeexplore.ieee.org/document/10298197/ |
| work_keys_str_mv | AT jonghyunkim efficientdatasetcollectionforconcretecrackdetectionwithspatialadaptivedataaugmentation AT junglee efficientdatasetcollectionforconcretecrackdetectionwithspatialadaptivedataaugmentation |