This repository hosts the MTC-UAV dataset associated with our work:

TasselNetV3: Explainable Plant Counting With Guided Upsampling and Background Suppression

Hao Lu¹, Liang Liu^1,, Ya-Nan Li², Xiao-Ming Zhao³, Xi-Qing Wang³, Zhi-Guo Cao¹

¹School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, China

²School of Computer Science and Engineering, Wuhan Institute of Technology, China

³Crop Functional Genomics and Molecular Breeding, College of Biological Science, China Agricultural University, China

IEEE Transactions on Geoscience and Remote Sensing, 2021

• over 400 maize cultivars
• 70,870 manually annotated instances
• 306 images with 5472x3648 image resolution
• 12.5-meter-height nadir imaging view

The MTC-UAV dataset can be downloaded from:

Google Drive (3.8 GB)

Suppose that the dataset is placed in the ./data folder, the path structure should look like:

$./data/maize_tassels_counting_uav_dataset
├──── images
│       ├──── DJI_0002.JPG
│       ├──── DJI_0007.JPG
│       ├──── ...
├──── labels
│       ├──── DJI_0002.csv
│       ├──── DJI_0007.csv
│       ├──── ...
├──── train.txt
├──── val.txt

• train.txt and val.txt record the train/val split used in the experiments.

If you find this work or code useful for your research, please cite:

@article{lu2021tasselnetv3,
  title={TasselNetV3: Explainable Plant Counting With Guided Upsampling and Background Suppression},
  author={Lu, Hao and Liu, Liang and Li, Ya-Nan and Zhao, Xiao-Ming and Wang, Xi-Qing and Cao, Zhiguo},
  journal={IEEE Transactions on Geoscience and Remote Sensing},
  year={2021}
}

This dataset is only for academic purposes. Contact Hao Lu ([email protected]) or Zhiguo Cao ([email protected]) for the commerial use of the model.