CROPUP

CROPUP: Historical products are all you need? An end-to-end cross-year crop map updating framework without the need for in situ samples

by Lei Lei, Xinyu Wang , Liangpei Zhang, Xin Hu and Yanfei Zhong

[] This is an official implementation of CROPUP in our RSE 2021 paper CROPUP: Historical products are all you need? An end-to-end cross-year crop map updating framework without the need for in situ samples .

Citation

If you use CROPUP in your research, please cite the following paper:

@article{LEI2024CROPUP,
title = {CROPUP: Historical products are all you need? An end-to-end cross-year crop map updating framework without the need for in situ samples},
journal = {Remote Sensing of Environment},
volume = {315},
pages = {114430},
year = {2024},
issn = {0034-4257},
doi = {https://doi.org/10.1016/j.rse.2024.114430},
url = {https://www.sciencedirect.com/science/article/pii/S0034425724004565},
author = {Lei Lei and Xinyu Wang and Liangpei Zhang and Xin Hu and Yanfei Zhong},
}

Getting started

Prepare environment

Prepare dataset

1. CDL download from GEE platform

2. Training and test dataset preparation

python data/cdldataset_rr.py

3. prepare congif file

Training and evaluation

python training_unite_cropup.py

Inferring for crop map

python inferring_unite.py