Binocular Video-Based Automatic Pixel-Level Crack Detection and Quantification Using Deep Convolutional Neural Networks for Concrete Structures
Crack detection and quantification play crucial roles in assessing the condition of concrete structures. Herein, a novel real-time crack detection and quantification method that leverages binocular vision and a lightweight deep learning model is proposed. In this methodology, the proposed method bas...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Buildings |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-5309/15/2/258 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832588861409067008 |
|---|---|
| author | Liqu Liu Bo Shen Shuchen Huang Runlin Liu Weizhang Liao Bin Wang Shuo Diao |
| author_facet | Liqu Liu Bo Shen Shuchen Huang Runlin Liu Weizhang Liao Bin Wang Shuo Diao |
| author_sort | Liqu Liu |
| collection | DOAJ |
| description | Crack detection and quantification play crucial roles in assessing the condition of concrete structures. Herein, a novel real-time crack detection and quantification method that leverages binocular vision and a lightweight deep learning model is proposed. In this methodology, the proposed method based on the following four modules is adopted: a lightweight classification algorithm, a high-precision segmentation algorithm, a semi-global block matching algorithm (SGBM), and a crack quantification technique. Based on the crack segmentation results, a framework is developed for quantitative analysis of the major geometric parameters, including crack length, crack width, and crack angle of orientation at the pixel level. Results indicate that, by incorporating channel attention and spatial attention mechanisms in the MBConv module, the detection accuracy of the improved EfficientNetV2 increased by 1.6% compared with the original EfficientNetV2. Results indicate that using the proposed quantification method can achieve low quantification errors of 2%, 4.5%, and 4% for the crack length, width, and angle of orientation, respectively. The proposed method can contribute to crack detection and quantification in practical use by being deployed on smart devices. |
| format | Article |
| id | doaj-art-c62790fa50ea4c5c8441a0edfae1ab95 |
| institution | Kabale University |
| issn | 2075-5309 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Buildings |
| spelling | doaj-art-c62790fa50ea4c5c8441a0edfae1ab952025-01-24T13:26:22ZengMDPI AGBuildings2075-53092025-01-0115225810.3390/buildings15020258Binocular Video-Based Automatic Pixel-Level Crack Detection and Quantification Using Deep Convolutional Neural Networks for Concrete StructuresLiqu Liu0Bo Shen1Shuchen Huang2Runlin Liu3Weizhang Liao4Bin Wang5Shuo Diao6State Key Laboratory of Building Safety and Built Environment, Beijing 100013, ChinaState Key Laboratory of Building Safety and Built Environment, Beijing 100013, ChinaBeijing Higher Institution Engineering Research Center of Structural Engineering and New Materials, Beijing University of Civil Engineering and Architecture, Beijing 100044, ChinaBeijing Higher Institution Engineering Research Center of Structural Engineering and New Materials, Beijing University of Civil Engineering and Architecture, Beijing 100044, ChinaBeijing Higher Institution Engineering Research Center of Structural Engineering and New Materials, Beijing University of Civil Engineering and Architecture, Beijing 100044, ChinaBeijing Higher Institution Engineering Research Center of Structural Engineering and New Materials, Beijing University of Civil Engineering and Architecture, Beijing 100044, ChinaState Key Laboratory of Building Safety and Built Environment, Beijing 100013, ChinaCrack detection and quantification play crucial roles in assessing the condition of concrete structures. Herein, a novel real-time crack detection and quantification method that leverages binocular vision and a lightweight deep learning model is proposed. In this methodology, the proposed method based on the following four modules is adopted: a lightweight classification algorithm, a high-precision segmentation algorithm, a semi-global block matching algorithm (SGBM), and a crack quantification technique. Based on the crack segmentation results, a framework is developed for quantitative analysis of the major geometric parameters, including crack length, crack width, and crack angle of orientation at the pixel level. Results indicate that, by incorporating channel attention and spatial attention mechanisms in the MBConv module, the detection accuracy of the improved EfficientNetV2 increased by 1.6% compared with the original EfficientNetV2. Results indicate that using the proposed quantification method can achieve low quantification errors of 2%, 4.5%, and 4% for the crack length, width, and angle of orientation, respectively. The proposed method can contribute to crack detection and quantification in practical use by being deployed on smart devices.https://www.mdpi.com/2075-5309/15/2/258concrete structurescrack detectionEfficientNetV2binocular visioncrack quantificationdeep learning |
| spellingShingle | Liqu Liu Bo Shen Shuchen Huang Runlin Liu Weizhang Liao Bin Wang Shuo Diao Binocular Video-Based Automatic Pixel-Level Crack Detection and Quantification Using Deep Convolutional Neural Networks for Concrete Structures Buildings concrete structures crack detection EfficientNetV2 binocular vision crack quantification deep learning |
| title | Binocular Video-Based Automatic Pixel-Level Crack Detection and Quantification Using Deep Convolutional Neural Networks for Concrete Structures |
| title_full | Binocular Video-Based Automatic Pixel-Level Crack Detection and Quantification Using Deep Convolutional Neural Networks for Concrete Structures |
| title_fullStr | Binocular Video-Based Automatic Pixel-Level Crack Detection and Quantification Using Deep Convolutional Neural Networks for Concrete Structures |
| title_full_unstemmed | Binocular Video-Based Automatic Pixel-Level Crack Detection and Quantification Using Deep Convolutional Neural Networks for Concrete Structures |
| title_short | Binocular Video-Based Automatic Pixel-Level Crack Detection and Quantification Using Deep Convolutional Neural Networks for Concrete Structures |
| title_sort | binocular video based automatic pixel level crack detection and quantification using deep convolutional neural networks for concrete structures |
| topic | concrete structures crack detection EfficientNetV2 binocular vision crack quantification deep learning |
| url | https://www.mdpi.com/2075-5309/15/2/258 |
| work_keys_str_mv | AT liquliu binocularvideobasedautomaticpixellevelcrackdetectionandquantificationusingdeepconvolutionalneuralnetworksforconcretestructures AT boshen binocularvideobasedautomaticpixellevelcrackdetectionandquantificationusingdeepconvolutionalneuralnetworksforconcretestructures AT shuchenhuang binocularvideobasedautomaticpixellevelcrackdetectionandquantificationusingdeepconvolutionalneuralnetworksforconcretestructures AT runlinliu binocularvideobasedautomaticpixellevelcrackdetectionandquantificationusingdeepconvolutionalneuralnetworksforconcretestructures AT weizhangliao binocularvideobasedautomaticpixellevelcrackdetectionandquantificationusingdeepconvolutionalneuralnetworksforconcretestructures AT binwang binocularvideobasedautomaticpixellevelcrackdetectionandquantificationusingdeepconvolutionalneuralnetworksforconcretestructures AT shuodiao binocularvideobasedautomaticpixellevelcrackdetectionandquantificationusingdeepconvolutionalneuralnetworksforconcretestructures |