# Pavementscapes

**Repository Path**: fengstyle/Pavementscapes

## Basic Information

- **Project Name**: Pavementscapes
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2026-06-09
- **Last Updated**: 2026-06-09

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Pavementscapes
Pavementscapes is a pixel-wise database for pavement surface distress segmentation. The dataset descriptions, including the statistical analysis and deep learning baselines, has been pre-printed in https://arxiv.org/abs/2208.00775.

Users can download the dataset by [google drive](https://drive.google.com/drive/folders/1h9Mis8s93TrRfzWQ4tThi8D_Ze0qS25U?usp=sharing).

The mask_maker2.ipynb shows a demo how to make the TFrecords files for the dataset and use this dataset to train and test a deep neural networks.

If having any problems, please feel free to contact Dr. Zheng Tong (tongzheng@seu.edu.cn).

Please use the below citation if you use our database, thank you.

--------------------------------

@misc{pavementscapes,

  doi = {10.48550/ARXIV.2208.00775},
  
  author = {Tong, Zheng and Ma, Tao and Huyan, Ju and Zhang, Weiguang},
  
  title = {Pavementscapes: a large-scale hierarchical image dataset for asphalt pavement damage segmentation},
  
  publisher = {arXiv},
  
  year = {2022},
  
  copyright = {arXiv.org perpetual, non-exclusive license}
  
}