RasterDataset: add control over resampling algorithm (#2015)

* RasterDataset: add control over resampling algorithm

* Fix type hints

* cubic -> bilinear

* Ruff: single quotes
This commit is contained in:
Adam J. Stewart 2024-05-13 17:00:45 +02:00 коммит произвёл GitHub
Родитель 5976bd15bf
Коммит 25fb9ccfb9
Не найден ключ, соответствующий данной подписи
Идентификатор ключа GPG: B5690EEEBB952194
3 изменённых файлов: 54 добавлений и 2 удалений

Просмотреть файл

@ -336,6 +336,10 @@
"\n", "\n",
"Defaults to float32 for `is_image == True` and long for `is_image == False`. This is what you want for 99% of datasets, but can be overridden for tasks like pixel-wise regression (where the target mask should be float32).\n", "Defaults to float32 for `is_image == True` and long for `is_image == False`. This is what you want for 99% of datasets, but can be overridden for tasks like pixel-wise regression (where the target mask should be float32).\n",
"\n", "\n",
"### `resampling`\n",
"\n",
"Defaults to bilinear for float Tensors and nearest for int Tensors. Can be overridden for custom resampling algorithms.\n",
"\n",
"### `separate_files`\n", "### `separate_files`\n",
"\n", "\n",
"If your data comes with each spectral band in a separate files, as is the case with Sentinel-2, use `separate_files = True`. If all spectral bands are stored in a single file, use `separate_files = False` instead.\n", "If your data comes with each spectral band in a separate files, as is the case with Sentinel-2, use `separate_files = True`. If all spectral bands are stored in a single file, use `separate_files = False` instead.\n",

Просмотреть файл

@ -5,12 +5,14 @@ import pickle
import sys import sys
from collections.abc import Iterable from collections.abc import Iterable
from pathlib import Path from pathlib import Path
from typing import Any
import pytest import pytest
import torch import torch
import torch.nn as nn import torch.nn as nn
from _pytest.fixtures import SubRequest from _pytest.fixtures import SubRequest
from rasterio.crs import CRS from rasterio.crs import CRS
from rasterio.enums import Resampling
from torch.utils.data import ConcatDataset from torch.utils.data import ConcatDataset
from torchgeo.datasets import ( from torchgeo.datasets import (
@ -49,6 +51,16 @@ class CustomGeoDataset(GeoDataset):
return {'index': bounds} return {'index': bounds}
class CustomRasterDataset(RasterDataset):
def __init__(self, dtype: torch.dtype, *args: Any, **kwargs: Any) -> None:
super().__init__(*args, **kwargs)
self._dtype = dtype
@property
def dtype(self) -> torch.dtype:
return self._dtype
class CustomVectorDataset(VectorDataset): class CustomVectorDataset(VectorDataset):
filename_glob = '*.geojson' filename_glob = '*.geojson'
date_format = '%Y' date_format = '%Y'
@ -274,6 +286,22 @@ class TestRasterDataset:
assert isinstance(x['image'], torch.Tensor) assert isinstance(x['image'], torch.Tensor)
assert x['image'].dtype == torch.float32 assert x['image'].dtype == torch.float32
@pytest.mark.parametrize('dtype', [torch.float, torch.double])
def test_resampling_float_dtype(self, dtype: torch.dtype) -> None:
paths = os.path.join('tests', 'data', 'raster', 'uint16')
ds = CustomRasterDataset(dtype, paths)
x = ds[ds.bounds]
assert x['image'].dtype == dtype
assert ds.resampling == Resampling.bilinear
@pytest.mark.parametrize('dtype', [torch.long, torch.bool])
def test_resampling_int_dtype(self, dtype: torch.dtype) -> None:
paths = os.path.join('tests', 'data', 'raster', 'uint16')
ds = CustomRasterDataset(dtype, paths)
x = ds[ds.bounds]
assert x['image'].dtype == dtype
assert ds.resampling == Resampling.nearest
def test_invalid_query(self, sentinel: Sentinel2) -> None: def test_invalid_query(self, sentinel: Sentinel2) -> None:
query = BoundingBox(0, 0, 0, 0, 0, 0) query = BoundingBox(0, 0, 0, 0, 0, 0)
with pytest.raises( with pytest.raises(

Просмотреть файл

@ -22,6 +22,7 @@ import rasterio.merge
import shapely import shapely
import torch import torch
from rasterio.crs import CRS from rasterio.crs import CRS
from rasterio.enums import Resampling
from rasterio.io import DatasetReader from rasterio.io import DatasetReader
from rasterio.vrt import WarpedVRT from rasterio.vrt import WarpedVRT
from rtree.index import Index, Property from rtree.index import Index, Property
@ -309,7 +310,7 @@ class GeoDataset(Dataset[dict[str, Any]], abc.ABC):
files |= set(glob.iglob(pathname, recursive=True)) files |= set(glob.iglob(pathname, recursive=True))
elif os.path.isfile(path) or path_is_vsi(path): elif os.path.isfile(path) or path_is_vsi(path):
files.add(path) files.add(path)
elif not hasattr(self, "download"): elif not hasattr(self, 'download'):
warnings.warn( warnings.warn(
f"Could not find any relevant files for provided path '{path}'. " f"Could not find any relevant files for provided path '{path}'. "
f'Path was ignored.', f'Path was ignored.',
@ -384,6 +385,23 @@ class RasterDataset(GeoDataset):
else: else:
return torch.long return torch.long
@property
def resampling(self) -> Resampling:
"""Resampling algorithm used when reading input files.
Defaults to bilinear for float dtypes and nearest for int dtypes.
Returns:
The resampling method to use.
.. versionadded:: 0.6
"""
# Based on torch.is_floating_point
if self.dtype in [torch.float64, torch.float32, torch.float16, torch.bfloat16]:
return Resampling.bilinear
else:
return Resampling.nearest
def __init__( def __init__(
self, self,
paths: str | Iterable[str] = 'data', paths: str | Iterable[str] = 'data',
@ -555,7 +573,9 @@ class RasterDataset(GeoDataset):
vrt_fhs = [self._load_warp_file(fp) for fp in filepaths] vrt_fhs = [self._load_warp_file(fp) for fp in filepaths]
bounds = (query.minx, query.miny, query.maxx, query.maxy) bounds = (query.minx, query.miny, query.maxx, query.maxy)
dest, _ = rasterio.merge.merge(vrt_fhs, bounds, self.res, indexes=band_indexes) dest, _ = rasterio.merge.merge(
vrt_fhs, bounds, self.res, indexes=band_indexes, resampling=self.resampling
)
# Use array_to_tensor since merge may return uint16/uint32 arrays. # Use array_to_tensor since merge may return uint16/uint32 arrays.
tensor = array_to_tensor(dest) tensor = array_to_tensor(dest)
return tensor return tensor