ying-fu / hfd100

A Large-Scale Hyperspectral Dataset for Flower Classification

This is HFD100 dataset and implementation of A Large-Scale Hyperspectral Dataset for Flower Classification in KBS(Knowledge-Based Systems), 2022, by Yongrong Zheng, Tao Zhang, Ying Fu.

---

Paper

---

HFD100 dataset

1. The data format used for the experiment is .mat，the division ratio of training dataset and test dataset is 3:2.
2. The number of channels for hyperspectral image is 31.
3. The size of HFD100 dataset is about 55 Gb, the download link is https://pan.baidu.com/s/1XKGafuIbFpI3V77LSeFbOA and the code is "epou".

Citation

If you use our dataset or code for research, please ensure that you cite our paper:

@article{zheng2022large,
  title={A large-scale hyperspectral dataset for flower classification},
  author={Zheng, Yongrong and Zhang, Tao and Fu, Ying},
  journal={Knowledge-Based Systems},
  volume={236},
  pages={107647--107658},
  year={2022},
  publisher={Elsevier}
}

hfd100's Issues

Want to obtain relevant documents

hello， Can you please publish the training-related train.h5 file or the related code, and then I can reproduce your base-line. thank you

HFD100 data

Hello, where can I get the HFD100 data set