Zobraziť v EDS

Rapid extraction of pavement aggregate gradation based on point clouds using deep learning networks

Uložené v:
Podrobná bibliografia
Názov: Rapid extraction of pavement aggregate gradation based on point clouds using deep learning networks
Autori: Chen, Siyu (author), Chen, Can (author), Ma, Tao (author), Han, Chengjia (author), Luo, Haoyuan (author), Wang, Siqi (author), Gao, Y. (author), Yang, Yaowen (author)
Prispievatelia: School of Civil and Environmental Engineering
Zdroj: Automation in Construction. 154:105023
Informácie o vydavateľovi: Elsevier BV, 2023.
Rok vydania: 2023
Predmety: Artificial neural networks, Civil engineering [Engineering], Asphalt Pavement, Multi-feature fusion, Aggregate gradation, Aggregate Gradation, 0211 other engineering and technologies, Asphalt pavement, 02 engineering and technology, Point clouds, 0201 civil engineering
Popis: Usage of asphalt mixture with poor gradation will most likely lead to pavement deficiency. There is a growing need for rapid and non-destructive methods to extract pavement aggregate gradation. In this study, a deep learning-based method that utilizes point clouds data for gradation extraction was proposed. Firstly, a data enhancement algorithm along with three data format conversion methods (aligned point cloud, voxel, and depth image) were proposed to preprocess the original collected point clouds. Subsequently, different neural network models were designed for each data format to extract gradation. Finally, a multi-feature fusion network was developed, which using extraction network as the backbone and additional auxiliary information. In the case study, the MAE loss of multi-feature fusion networks with PointNet, Vox-ResNet34 and GoogLeNet-v4 as the backbone respectively achieved 0.202, 0.142 and 0.046 on the test set, which means an estimation accuracy of more than 95% for the pavement aggregate gradation. AI Singapore National Research Foundation (NRF) Submitted/Accepted version This paper is part of the research work of National Key Research and Development Project of China (Grant No. 2021YFB2600601, 2021YFB2600600). The authors would like to acknowledge the financial support provided by the National Natural Science Foundation of China (Grant No. 51922030), Natural Science Foundation of Jiangsu (Grant No. BK20220845), “the Fundamental Research Funds for the Central Universities” (Grant No. 2242022R10019). This research is also supported by the National Research Foundation, Singapore under its AI Singapore Programme (AISG Award No: AISG2-TC-2021-001).
Druh dokumentu: Article
Popis súboru: application/pdf
Jazyk: English
ISSN: 0926-5805
DOI: 10.1016/j.autcon.2023.105023
Prístupová URL adresa: https://hdl.handle.net/10356/170908
http://resolver.tudelft.nl/uuid:feb5e4f9-770a-447f-bd41-571e992ad79a
Rights: Elsevier TDM
Prístupové číslo: edsair.doi.dedup.....c3b9fb76334d3d5d50a77391ea4d2118
Databáza: OpenAIRE
Popis
Abstrakt:Usage of asphalt mixture with poor gradation will most likely lead to pavement deficiency. There is a growing need for rapid and non-destructive methods to extract pavement aggregate gradation. In this study, a deep learning-based method that utilizes point clouds data for gradation extraction was proposed. Firstly, a data enhancement algorithm along with three data format conversion methods (aligned point cloud, voxel, and depth image) were proposed to preprocess the original collected point clouds. Subsequently, different neural network models were designed for each data format to extract gradation. Finally, a multi-feature fusion network was developed, which using extraction network as the backbone and additional auxiliary information. In the case study, the MAE loss of multi-feature fusion networks with PointNet, Vox-ResNet34 and GoogLeNet-v4 as the backbone respectively achieved 0.202, 0.142 and 0.046 on the test set, which means an estimation accuracy of more than 95% for the pavement aggregate gradation. AI Singapore National Research Foundation (NRF) Submitted/Accepted version This paper is part of the research work of National Key Research and Development Project of China (Grant No. 2021YFB2600601, 2021YFB2600600). The authors would like to acknowledge the financial support provided by the National Natural Science Foundation of China (Grant No. 51922030), Natural Science Foundation of Jiangsu (Grant No. BK20220845), “the Fundamental Research Funds for the Central Universities” (Grant No. 2242022R10019). This research is also supported by the National Research Foundation, Singapore under its AI Singapore Programme (AISG Award No: AISG2-TC-2021-001).
ISSN:09265805
DOI:10.1016/j.autcon.2023.105023