Airphen: add new dataset

docs/api/datasets.rst

@@ -23,6 +23,11 @@ Aboveground Woody Biomass
 
 .. autoclass:: AbovegroundLiveWoodyBiomassDensity
 
+Airphen
+^^^^^^^
+
+.. autoclass:: Airphen
+
 Aster Global DEM
 ^^^^^^^^^^^^^^^^
 

docs/api/geo_datasets.csv

@@ -1,5 +1,6 @@
 Dataset,Type,Source,License,Size (px),Resolution (m)
 `Aboveground Woody Biomass`_,Masks,"Landsat, LiDAR","CC-BY-4.0","40,000x40,000",30
+`Airphen`_,Imagery,Airphen,-,"1,280x960",0.047--0.09
 `Aster Global DEM`_,Masks,Aster,"public domain","3,601x3,601",30
 `Canadian Building Footprints`_,Geometries,Bing Imagery,"ODbL-1.0",-,-
 `Chesapeake Land Cover`_,"Imagery, Masks",NAIP,"CC-BY-4.0",-,1

tests/data/airphen/data.py

@@ -0,0 +1,39 @@
+#!/usr/bin/env python3
+
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+import numpy as np
+import rasterio
+from rasterio import Affine
+from rasterio.crs import CRS
+
+SIZE = 36
+
+np.random.seed(0)
+
+
+profile = {
+    "driver": "GTiff",
+    "dtype": "uint16",
+    "width": SIZE,
+    "height": SIZE,
+    "count": 6,
+    "crs": CRS.from_epsg(4326),
+    "transform": Affine(
+        4.497249999999613e-07,
+        0.0,
+        12.567765446921205,
+        0.0,
+        -4.4972499999996745e-07,
+        47.42974580435403,
+    ),
+}
+
+Z = np.random.randint(
+    np.iinfo(profile["dtype"]).max, size=(SIZE, SIZE), dtype=profile["dtype"]
+)
+
+with rasterio.open("zoneA_B_R_NIR.tif", "w", **profile) as src:
+    for i in range(profile["count"]):
+        src.write(Z, i + 1)

tests/datasets/test_airphen.py

@@ -0,0 +1,71 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+import os
+from pathlib import Path
+
+import matplotlib.pyplot as plt
+import pytest
+import torch
+import torch.nn as nn
+from rasterio.crs import CRS
+
+from torchgeo.datasets import (
+    Airphen,
+    BoundingBox,
+    DatasetNotFoundError,
+    IntersectionDataset,
+    RGBBandsMissingError,
+    UnionDataset,
+)
+
+
+class TestAirphen:
+    @pytest.fixture
+    def dataset(self) -> Airphen:
+        paths = os.path.join("tests", "data", "airphen")
+        bands = ["B1", "B3", "B4"]
+        transforms = nn.Identity()
+        return Airphen(paths, bands=bands, transforms=transforms)
+
+    def test_len(self, dataset: Airphen) -> None:
+        assert len(dataset) == 1
+
+    def test_getitem(self, dataset: Airphen) -> None:
+        x = dataset[dataset.bounds]
+        assert isinstance(x, dict)
+        assert isinstance(x["crs"], CRS)
+        assert isinstance(x["image"], torch.Tensor)
+
+    def test_and(self, dataset: Airphen) -> None:
+        ds = dataset & dataset
+        assert isinstance(ds, IntersectionDataset)
+
+    def test_or(self, dataset: Airphen) -> None:
+        ds = dataset | dataset
+        assert isinstance(ds, UnionDataset)
+
+    def test_plot(self, dataset: Airphen) -> None:
+        x = dataset[dataset.bounds]
+        dataset.plot(x, suptitle="Test")
+        plt.close()
+
+    def test_no_data(self, tmp_path: Path) -> None:
+        with pytest.raises(DatasetNotFoundError, match="Dataset not found"):
+            Airphen(str(tmp_path))
+
+    def test_invalid_query(self, dataset: Airphen) -> None:
+        query = BoundingBox(0, 0, 0, 0, 0, 0)
+        with pytest.raises(
+            IndexError, match="query: .* not found in index with bounds:"
+        ):
+            dataset[query]
+
+    def test_plot_wrong_bands(self, dataset: Airphen) -> None:
+        bands = ("B1", "B2", "B3")
+        ds = Airphen(dataset.paths, bands=bands)
+        x = dataset[dataset.bounds]
+        with pytest.raises(
+            RGBBandsMissingError, match="Dataset does not contain some of the RGB bands"
+        ):
+            ds.plot(x)

torchgeo/datasets/__init__.py

@@ -5,6 +5,7 @@
 
 from .advance import ADVANCE
 from .agb_live_woody_density import AbovegroundLiveWoodyBiomassDensity
+from .airphen import Airphen
 from .astergdem import AsterGDEM
 from .benin_cashews import BeninSmallHolderCashews
 from .bigearthnet import BigEarthNet
@@ -135,6 +136,7 @@
 __all__ = (
     # GeoDataset
     "AbovegroundLiveWoodyBiomassDensity",
+    "Airphen",
     "AsterGDEM",
     "CanadianBuildingFootprints",
     "CDL",

torchgeo/datasets/airphen.py

@@ -0,0 +1,88 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+"""Airphen dataset."""
+
+from typing import Any, Optional
+
+import matplotlib.pyplot as plt
+from matplotlib.figure import Figure
+
+from .geo import RasterDataset
+from .utils import RGBBandsMissingError, percentile_normalization
+
+
+class Airphen(RasterDataset):
+    """Airphen dataset.
+
+    `Airphen <https://ideol.sakura.ne.jp/img/20170123_HiphenAirphenKeyfeatures.pdf>`__
+    is a multispectral scientific camera developed by agronomists and photonics
+    engineers at `Hiphen <https://www.hiphen-plant.com/>`_ to match plant measurements
+    needs and constraints. Its high flexibility, ease of use and radiometric quality
+    give you a wide range of opportunities to develop your customized applications.
+
+    Main characteristics:
+
+    * 6 Synchronized global shutter sensors
+    * Sensor resolution 1280 x 960 pixels
+    * Data format (.tiff, 12 bit)
+    * SD card storage
+    * Metadata information: Exif and XMP
+    * Internal or external GPS
+    * Synchronization with different sensors (TIR, RGB, others)
+
+    Key features:
+
+    * Customized sensing configurations
+    * Easy integration
+    * Full control and configuration
+    * Powerful analysis with a specific Agisoft module
+
+    .. versionadded:: 0.6
+    """
+
+    # Each camera measures a custom set of spectral bands chosen at purchase time.
+    # Hiphen offers 8 bands to choose from, sorted from short to long wavelength.
+    all_bands = ["B1", "B2", "B3", "B4", "B5", "B6", "B7", "B8"]
+    rgb_bands = ["B4", "B3", "B1"]
+
+    def plot(
+        self,
+        sample: dict[str, Any],
+        show_titles: bool = True,
+        suptitle: Optional[str] = None,
+    ) -> Figure:
+        """Plot a sample from the dataset.
+
+        Args:
+            sample: a sample returned by :meth:`RasterDataset.__getitem__`
+            show_titles: flag indicating whether to show titles above each panel
+            suptitle: optional string to use as a suptitle
+
+        Returns:
+            a matplotlib Figure with the rendered sample
+
+        Raises:
+            RGBBandsMissingError: If *bands* does not include all RGB bands.
+        """
+        rgb_indices = []
+        for band in self.rgb_bands:
+            if band in self.bands:
+                rgb_indices.append(self.bands.index(band))
+            else:
+                raise RGBBandsMissingError()
+
+        image = sample["image"][rgb_indices].permute(1, 2, 0).float()
+        image = percentile_normalization(image, axis=(0, 1))
+
+        fig, ax = plt.subplots(1, 1, figsize=(4, 4))
+        ax.imshow(image)
+        ax.axis("off")
+
+        if show_titles:
+            ax.set_title("Image")
+
+        if suptitle is not None:
+            plt.suptitle(suptitle)
+
+        return fig