S4A.py

# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.


from itertools import product
import numpy as np
import xarray as xr
import dask
import netCDF4
import datasets
from pathlib import Path


_CITATION = """\
@ARTICLE{
  9749916,
  author={Sykas, Dimitrios and Sdraka, Maria and Zografakis, Dimitrios and Papoutsis, Ioannis},
  journal={IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
  title={A Sentinel-2 multi-year, multi-country benchmark dataset for crop classification and segmentation with deep learning},
  year={2022},
  doi={10.1109/JSTARS.2022.3164771}
}
"""

_DESCRIPTION = """\
Sen4AgriNet is a Sentinel-2 based time series multi country benchmark dataset, tailored for 
agricultural monitoring applications with Machine and Deep Learning. It is annotated from 
farmer declarations collected via the Land Parcel Identification System (LPIS) for harmonizing 
country wide labels. These declarations have only recently been made available as open data, 
allowing for the first time the labelling of satellite imagery from ground truth data. 
We proceed to propose and standardise a new crop type taxonomy across Europe that address 
Common Agriculture Policy (CAP) needs, based on the Food and Agriculture Organization (FAO) 
Indicative Crop Classification scheme. Sen4AgriNet is the only multi-country, multi-year dataset 
that includes all spectral information. It is constructed to cover the period 2016-2020 for 
Catalonia and France, while it can be extended to include additional countries.
"""

_HOMEPAGE = "https://www.sen4agrinet.space.noa.gr/"

_LICENSE = "MIT License"

_URL = 'https://huggingface.co/datasets/paren8esis/S4A/resolve/main/data'

# The tiles of Catalonia
CAT_TILES = ['31TBF', '31TCF', '31TCG', '31TDF', '31TDG']

# The tiles of France
FR_TILES = ['31TCJ', '31TDK', '31TCL', '31TDM', '31UCP', '31UDR']

# The maximum indices for each patch 
PATCH_IDX = {
    '2019': {
        '31TBF': [29, 29],
        '31TCF': [29, 27],
        '31TCG': [29, 29],
        '31TDF': [15, 9], 
        '31TDG': [29, 29],
        '31TCJ': [29, 29], 
        '31TDK': [29, 29], 
        '31TCL': [29, 29], 
        '31TDM': [29, 29], 
        '31UCP': [29, 29], 
        '31UDR': [29, 29]
    },
    '2020': {
        '31TBF': [29, 29],
        '31TCF': [29, 27],
        '31TCG': [29, 29],
        '31TDF': [15, 9], 
        '31TDG': [29, 29]
    }
}


class S4A(datasets.GeneratorBasedBuilder):
    VERSION = datasets.Version("0.0.1")

    BUILDER_CONFIGS = [
        datasets.BuilderConfig(name="complete", version=VERSION, description="All Sen4AgriNet data."),
        datasets.BuilderConfig(name="tiny", version=VERSION, description="Just three samples for testing."),
        datasets.BuilderConfig(name="cat_2019", version=VERSION, description="Sen4AgriNet data for Catalonia 2019."),
    ]

    DEFAULT_CONFIG_NAME = "complete"

    def _info(self):
        features = datasets.Features(
            {
                "patch_full_name": datasets.Value("string"),
                "patch_year": datasets.Value("string"),
                "patch_name": datasets.Value("string"),
                "patch_country_code": datasets.Value("string"),
                "patch_tile": datasets.Value("string"),
                "B01": datasets.Array3D(shape=(None, 61, 61), dtype="uint16"),
                "B02": datasets.Array3D(shape=(None, 366, 366), dtype="uint16"),
                "B03": datasets.Array3D(shape=(None, 366, 366), dtype="uint16"),
                "B04": datasets.Array3D(shape=(None, 366, 366), dtype="uint16"),
                "B05": datasets.Array3D(shape=(None, 183, 183), dtype="uint16"),
                "B06": datasets.Array3D(shape=(None, 183, 183), dtype="uint16"),
                "B07": datasets.Array3D(shape=(None, 183, 183), dtype="uint16"),
                "B08": datasets.Array3D(shape=(None, 366, 366), dtype="uint16"),
                "B09": datasets.Array3D(shape=(None, 61, 61), dtype="uint16"),
                "B10": datasets.Array3D(shape=(None, 61, 61), dtype="uint16"),
                "B11": datasets.Array3D(shape=(None, 183, 183), dtype="uint16"),
                "B12": datasets.Array3D(shape=(None, 183, 183), dtype="uint16"),
                "B8A": datasets.Array3D(shape=(None, 183, 183), dtype="uint16"),
                "labels": datasets.Array2D(shape=(366, 366), dtype="uint32"),
                "parcels": datasets.Array2D(shape=(366, 366), dtype="uint32"),
                "timestamp": datasets.Sequence(datasets.Value("timestamp[ns]"))
            }
        )
    
        return datasets.DatasetInfo(
            # This is the description that will appear on the datasets page.
            description=_DESCRIPTION,
            features=features,
            homepage=_HOMEPAGE,
            license=_LICENSE,
            citation=_CITATION,
        )


    def _split_generators(self, dl_manager):
        root_paths = []
        if self.config.name == "complete":
            for year, tile in list(product(['2019'], FR_TILES)) + list(product(['2019', '2020'], CAT_TILES)):
                x, y = PATCH_IDX[year][tile]
                for x_i in range(x + 1):
                    for y_i in range(y + 1):
                        downloaded_paths = dl_manager.download(_URL + f'/{year}' + f'/{tile}' + f'/{year}_{tile}_patch_{str(x_i).zfill(2)}_{str(y_i).zfill(2)}.nc')
                        root_paths.append(downloaded_paths)

            return [
                datasets.SplitGenerator(
                    name='complete',
                    # These kwargs will be passed to _generate_examples
                    gen_kwargs={
                        "root_paths": root_paths,
                    },
                ),
            ]
        elif self.config.name == 'tiny':
            selected_files = ['2019_31TCF_patch_00_11.nc', '2019_31UCP_patch_01_19.nc', '2020_31TDG_patch_11_17.nc']

            for file in selected_files:
                year, tile = file.split('_')[:2]

                downloaded_paths = dl_manager.download(_URL + f'/{year}' + f'/{tile}' + f'/{file}')
                root_paths.append(downloaded_paths)

            return [
                datasets.SplitGenerator(
                    name='tiny',
                    # These kwargs will be passed to _generate_examples
                    gen_kwargs={
                        "root_paths": root_paths,
                    },
                ),
            ]


    # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
    def _generate_examples(self, root_paths):
        for file in root_paths:
            data = xr.open_dataset(file, chunks=-1, engine='netcdf4')

            res = {
                "patch_full_name": data.patch_full_name,
                "patch_year": data.patch_year,
                "patch_name": data.patch_name,
                "patch_country_code": data.patch_country_code,
                "patch_tile": data.patch_tile
            }

            time_recorded = False

            for variable in ['B01', 'B02', 'B03', 'B04', 'B05', 'B06', 'B07', 'B08', 'B09', 'B10', 'B11', 'B12', 'B8A', 'labels', 'parcels']:
                v = xr.open_dataset(file, chunks=-1, engine='netcdf4', group=variable)
                if not time_recorded:
                    res['timestamp'] = (v.time.values.astype(np.int64) // 10 ** 9).tolist()
                    time_recorded = True

                res[variable] = getattr(v, variable).values

            key = res['patch_full_name']
            yield key, res