MS-FPN-Based Pavement Defect Identification Algorithm
False or missed detection may happen in the process of pavement defect detection due to cracks with different shapes and sizes and interference from complex pavement background. To solve this problem, we proposed a Cascade R-CNN detection method based on the MS-Feature Pyramid Network (MS-FPN). Firs...
| Principais autores: | , , , |
|---|---|
| Formato: | Artigo |
| Idioma: | English |
| Publicado em: |
IEEE
2023-01-01
|
| coleção: | IEEE Access |
| Assuntos: | |
| Acesso em linha: | https://ieeexplore.ieee.org/document/10304121/ |
| _version_ | 1827763162114424832 |
|---|---|
| author | Lei Chen Shiming An Shuang Zhao Guandian Li |
| author_facet | Lei Chen Shiming An Shuang Zhao Guandian Li |
| author_sort | Lei Chen |
| collection | DOAJ |
| description | False or missed detection may happen in the process of pavement defect detection due to cracks with different shapes and sizes and interference from complex pavement background. To solve this problem, we proposed a Cascade R-CNN detection method based on the MS-Feature Pyramid Network (MS-FPN). First, we introduced a deformable convolution module in the backbone network ResNet101, so that it can adaptively change depending on the pavement defect. Second, we utilized the MS-FPN for the cross-scale bi-directional fusion of feature maps output by the backbone network,in which the multi-branch hybrid dilated convolution (MCE) generates feature maps with multi-scale receptive fields while expanding the receptive field. The dual-channel attention fusion algorithm (ST-A) was used to improve the identification between the background and the object, so that more attention was paid to the location and features of the pavement defect object. The improved Cascade R-CNN can better adapt to the detection of various pavement defects. On the open-source dataset and self-built dataset, the detection precision has improved by 5.02% and 5% respectively compared to that of the original Cascade R-CNN. |
| first_indexed | 2024-03-11T10:48:15Z |
| format | Article |
| id | doaj.art-a5cdce96e9f74526a3fe2dfb81459565 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-11T10:48:15Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-a5cdce96e9f74526a3fe2dfb814595652023-11-14T00:00:35ZengIEEEIEEE Access2169-35362023-01-011112479712480710.1109/ACCESS.2023.332925010304121MS-FPN-Based Pavement Defect Identification AlgorithmLei Chen0Shiming An1https://orcid.org/0009-0004-1498-3412Shuang Zhao2Guandian Li3School of Electronic Information Engineering, Changchun University of Science and Technology, Changchun, ChinaSchool of Electronic Information Engineering, Changchun University of Science and Technology, Changchun, ChinaSchool of Electronic Information Engineering, Changchun University of Science and Technology, Changchun, ChinaSchool of Electronic Information Engineering, Changchun University of Science and Technology, Changchun, ChinaFalse or missed detection may happen in the process of pavement defect detection due to cracks with different shapes and sizes and interference from complex pavement background. To solve this problem, we proposed a Cascade R-CNN detection method based on the MS-Feature Pyramid Network (MS-FPN). First, we introduced a deformable convolution module in the backbone network ResNet101, so that it can adaptively change depending on the pavement defect. Second, we utilized the MS-FPN for the cross-scale bi-directional fusion of feature maps output by the backbone network,in which the multi-branch hybrid dilated convolution (MCE) generates feature maps with multi-scale receptive fields while expanding the receptive field. The dual-channel attention fusion algorithm (ST-A) was used to improve the identification between the background and the object, so that more attention was paid to the location and features of the pavement defect object. The improved Cascade R-CNN can better adapt to the detection of various pavement defects. On the open-source dataset and self-built dataset, the detection precision has improved by 5.02% and 5% respectively compared to that of the original Cascade R-CNN.https://ieeexplore.ieee.org/document/10304121/Pavement defectdeformable convolutiondual-channel attention fusionmulti-branch dilated convolutionMS-FPN |
| spellingShingle | Lei Chen Shiming An Shuang Zhao Guandian Li MS-FPN-Based Pavement Defect Identification Algorithm IEEE Access Pavement defect deformable convolution dual-channel attention fusion multi-branch dilated convolution MS-FPN |
| title | MS-FPN-Based Pavement Defect Identification Algorithm |
| title_full | MS-FPN-Based Pavement Defect Identification Algorithm |
| title_fullStr | MS-FPN-Based Pavement Defect Identification Algorithm |
| title_full_unstemmed | MS-FPN-Based Pavement Defect Identification Algorithm |
| title_short | MS-FPN-Based Pavement Defect Identification Algorithm |
| title_sort | ms fpn based pavement defect identification algorithm |
| topic | Pavement defect deformable convolution dual-channel attention fusion multi-branch dilated convolution MS-FPN |
| url | https://ieeexplore.ieee.org/document/10304121/ |
| work_keys_str_mv | AT leichen msfpnbasedpavementdefectidentificationalgorithm AT shimingan msfpnbasedpavementdefectidentificationalgorithm AT shuangzhao msfpnbasedpavementdefectidentificationalgorithm AT guandianli msfpnbasedpavementdefectidentificationalgorithm |