Binary image acquisition and texture parameter calculation of asphalt pavement based on a U-Net model

Fengwei An, Haoran Jiang, Yulong Zhao, Lei Fang, Shujing Dong, Wenpeng Du, Guohua Wu, Wei Liu, Zhenglong Lv and Hao Liang

PLOS ONE, 2026, vol. 21, issue 7, 1-1

Abstract: Current methods for detecting and evaluating pavement skid resistance vary widely, yet each has its own scope of applicability and inherent limitations. Therefore, this paper proposes a method based on U-Net model segmentation to obtain binary images of asphalt pavement surfaces, enabling precise calculation of pavement texture distribution parameters. A dataset required for model training and testing was constructed, and preprocessing was performed on forward-collected asphalt pavement images to eliminate noise interference. Based on the established three-dimensional asphalt pavement model, reverse engineering is employed to obtain binary images of the pavement surface as target images for model training. A U-Net semantic segmentation model is constructed to train the segmentation of asphalt pavement images, distinguishing between the aggregate and void components of the pavement. The feature parameters of the binary images generated by U-Net segmentation are calculated and correlated with the average texture depth measured using the sand patch method. Results indicate that the U-Net model achieved an F1-score of 0.7214 on the validation set, demonstrating satisfactory segmentation performance for subsequent texture parameter extraction. The correlation coefficient R² between the mean texture depth (MTD) calculated from the binary image and the sand patch method reached 0.85, while the correlation coefficient between the fractal dimension and MTD-1 was 0.86. Both correlations were statistically significant at the 95% confidence level. The proposed method can provide effective technical support for texture-based skid resistance evaluation of asphalt pavements.

Date: 2026
References: Add references at CitEc
Citations:

Downloads: (external link)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0354127 (text/html)
https://journals.plos.org/plosone/article/file?id= ... 54127&type=printable (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:plo:pone00:0354127

DOI: 10.1371/journal.pone.0354127

Access Statistics for this article

More articles in PLOS ONE from Public Library of Science
Bibliographic data for series maintained by plosone ().