Concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimization
Abstract Concrete frameworks require strong structural integrity to ensure their durability and performance. However, they are disposed to develop cracks, which can compromise their overall quality. This research presents an innovative crack diagnosis algorithm for concrete structures that utilizes...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-02-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-89250-3 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850067574504554496 |
|---|---|
| author | Yongqing Lin Mehdi Ahmadi Khalid A. Alnowibet Fawzy A. Bukhari |
| author_facet | Yongqing Lin Mehdi Ahmadi Khalid A. Alnowibet Fawzy A. Bukhari |
| author_sort | Yongqing Lin |
| collection | DOAJ |
| description | Abstract Concrete frameworks require strong structural integrity to ensure their durability and performance. However, they are disposed to develop cracks, which can compromise their overall quality. This research presents an innovative crack diagnosis algorithm for concrete structures that utilizes an optimized Deep Neural Network (DNN) called the Ridgelet Neural Network (RNN). The RNN model was then adjusted with a new advanced version of the Human Evolutionary Optimization (AHEO) algorithm that is introduced in this study. The AHEO as a new method combines human intelligence and evolutionary principles to optimize the RNN model. To train the model, an image dataset has been used, consisting of labeled images categorized as either “cracks” or “no-cracks”. The AHEO algorithm has been employed to refine the network’s weights, adjust the output layer for binary classification, and enhance the dataset through stochastic rotational augmentation. The effectiveness of the RNN/AHEO model was evaluated using various metrics and compared to existing methods. The model’s performance is evaluated by metrics such as accuracy, precision, recall, and F1-score, and is compared to existing methods including CNN, CrackUnet, R-CNN, DCNN, and U-Net, achieving an accuracy of 99.665% and an F1-score of 99.035%. The results demonstrated that the RNN/AHEO model outperformed other approaches in detecting concrete cracks. This innovative solution provides a robust method for maintaining the structural integrity of concrete frameworks. |
| format | Article |
| id | doaj-art-9ca0b25f1b5d4c2a80a65fe7b6a4e320 |
| institution | DOAJ |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-9ca0b25f1b5d4c2a80a65fe7b6a4e3202025-08-20T02:48:16ZengNature PortfolioScientific Reports2045-23222025-02-0115111610.1038/s41598-025-89250-3Concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimizationYongqing Lin0Mehdi Ahmadi1Khalid A. Alnowibet2Fawzy A. Bukhari3Beijing Jiaotong Vocational Technical CollegeAnkara Yıldırım Beyazıt University (AYBU)Statistics and Operations Research Department, College of Science, King Saud UniversityStatistics and Operations Research Department, College of Science, King Saud UniversityAbstract Concrete frameworks require strong structural integrity to ensure their durability and performance. However, they are disposed to develop cracks, which can compromise their overall quality. This research presents an innovative crack diagnosis algorithm for concrete structures that utilizes an optimized Deep Neural Network (DNN) called the Ridgelet Neural Network (RNN). The RNN model was then adjusted with a new advanced version of the Human Evolutionary Optimization (AHEO) algorithm that is introduced in this study. The AHEO as a new method combines human intelligence and evolutionary principles to optimize the RNN model. To train the model, an image dataset has been used, consisting of labeled images categorized as either “cracks” or “no-cracks”. The AHEO algorithm has been employed to refine the network’s weights, adjust the output layer for binary classification, and enhance the dataset through stochastic rotational augmentation. The effectiveness of the RNN/AHEO model was evaluated using various metrics and compared to existing methods. The model’s performance is evaluated by metrics such as accuracy, precision, recall, and F1-score, and is compared to existing methods including CNN, CrackUnet, R-CNN, DCNN, and U-Net, achieving an accuracy of 99.665% and an F1-score of 99.035%. The results demonstrated that the RNN/AHEO model outperformed other approaches in detecting concrete cracks. This innovative solution provides a robust method for maintaining the structural integrity of concrete frameworks.https://doi.org/10.1038/s41598-025-89250-3ConcreteCrack detectionRidgelet neural networkDeep neural networkAdvanced human evolutionary optimizationStructural integrity |
| spellingShingle | Yongqing Lin Mehdi Ahmadi Khalid A. Alnowibet Fawzy A. Bukhari Concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimization Scientific Reports Concrete Crack detection Ridgelet neural network Deep neural network Advanced human evolutionary optimization Structural integrity |
| title | Concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimization |
| title_full | Concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimization |
| title_fullStr | Concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimization |
| title_full_unstemmed | Concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimization |
| title_short | Concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimization |
| title_sort | concrete crack detection using ridgelet neural network optimized by advanced human evolutionary optimization |
| topic | Concrete Crack detection Ridgelet neural network Deep neural network Advanced human evolutionary optimization Structural integrity |
| url | https://doi.org/10.1038/s41598-025-89250-3 |
| work_keys_str_mv | AT yongqinglin concretecrackdetectionusingridgeletneuralnetworkoptimizedbyadvancedhumanevolutionaryoptimization AT mehdiahmadi concretecrackdetectionusingridgeletneuralnetworkoptimizedbyadvancedhumanevolutionaryoptimization AT khalidaalnowibet concretecrackdetectionusingridgeletneuralnetworkoptimizedbyadvancedhumanevolutionaryoptimization AT fawzyabukhari concretecrackdetectionusingridgeletneuralnetworkoptimizedbyadvancedhumanevolutionaryoptimization |