This repository implements SFC²Net proposed in the work:

High-Throughput Rice Density Estimation from Transplantation to Tillering Stages Using Deep Networks

Liang Liu, Hao Lu, Yanan Li, and Zhiguo Cao

Plant Phenomics, 2020

The code has been tested on Python 3.7.3 and PyTorch 1.3.1. Please follow the official instructions to configure your environment. See other required packages in requirements.txt.

• Download the Rice Plant Counting (RPC) test dataset from: BaiduYun (1.6 GB) (code: amg2)
• Unzip the dataset and move it into the ./data folder, the path structure should look like this:

$./data/rice_plant_counting_datasets
├──── images
├──── label_mat
├──── train.txt
├──── test.txt

Pre-trained Model on RPC dataset

• Download the model from: BaiduYun (26.2 MB) (code: nhxk) or OneDrive (26.2 MB)
• Move the model into the folder, and the path structure should be:

$./snapshots/rice/sfc2net
├──── model_best.pth.tar

Evaluation

python hltest.py

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

@article{liu2020high,
  title={High-Throughput Rice Density Estimation from Transplantation to Tillering Stages Using Deep Network},
  author={Liu, Liang and Lu, Hao and Li, Yanan and Cao, Zhiguo},
  journal={Plant Phenomics},
  year={2020}
}

The code and data are only for non-commercial purposes. Copyrights reserved.

The training set of the RPC dataset is made available upon request. Contact: Hao Lu ([email protected])