Build governed Belgian training corpus pipeline

2026-07-26 22:14:28 +02:00
parent 6323885a8d
commit 68d3fa34e3
5 changed files with 447 additions and 2 deletions
@@ -0,0 +1,71 @@
+from __future__ import annotations
+
+import importlib.util
+import json
+from pathlib import Path
+
+import numpy as np
+import rasterio
+from rasterio.transform import from_origin
+
+ROOT = Path(__file__).resolve().parents[2]
+SCRIPT = ROOT / "scripts" / "normalize_belgium_building_labels.py"
+SPEC = importlib.util.spec_from_file_location("normalize_buildings", SCRIPT)
+assert SPEC and SPEC.loader
+module = importlib.util.module_from_spec(SPEC)
+SPEC.loader.exec_module(module)
+
+
+def test_normalizer_retains_native_identity_and_records_rejections(tmp_path: Path) -> None:
+    raster_path = tmp_path / "image.tif"
+    with rasterio.open(
+        raster_path,
+        "w",
+        driver="GTiff",
+        width=100,
+        height=100,
+        count=3,
+        dtype="uint8",
+        crs="EPSG:4326",
+        transform=from_origin(4.0, 51.0, 0.001, 0.001),
+    ) as dataset:
+        dataset.write(np.zeros((3, 100, 100), dtype="uint8"))
+    valid = {
+        "type": "Feature",
+        "id": "native-1",
+        "properties": {"TYPE": "main building"},
+        "geometry": {"type": "Polygon", "coordinates": [[[4.01, 50.99], [4.02, 50.99], [4.02, 50.98], [4.01, 50.98], [4.01, 50.99]]]},
+    }
+    duplicate = json.loads(json.dumps(valid))
+    duplicate["id"] = "native-2"
+    canopy = json.loads(json.dumps(valid))
+    canopy["id"] = "native-3"
+    canopy["properties"]["TYPE"] = "canopy"
+    tiny = json.loads(json.dumps(valid))
+    tiny["id"] = "native-4"
+    tiny["geometry"] = {"type": "Polygon", "coordinates": [[[4.03, 50.97], [4.031, 50.97], [4.031, 50.969], [4.03, 50.969], [4.03, 50.97]]]}
+    reference_path = tmp_path / "reference.geojson"
+    reference_path.write_text(json.dumps({"type": "FeatureCollection", "features": [valid, duplicate, canopy, tiny]}), encoding="utf-8")
+
+    normalized, audit = module.normalize(
+        reference_path=reference_path,
+        raster_path=raster_path,
+        source_name="urbis",
+        min_label_px=3,
+        imagery_observed_at="2026-01-01T00:00:00Z",
+        reference_observed_at="2025-12-01T00:00:00Z",
+    )
+
+    assert len(normalized["features"]) == 1
+    properties = normalized["features"][0]["properties"]
+    assert properties["canonical_class"] == "building"
+    assert properties["source_name"] == "urbis"
+    assert properties["source_feature_id"] == "native-1"
+    assert properties["source_class"] == "main building"
+    assert audit["decision_counts"] == {
+        "accepted": 1,
+        "below_resolvable_pixel_size": 1,
+        "duplicate_geometry": 1,
+        "excluded_canopy": 1,
+    }
+    assert audit["temporal_mismatch_days"] == 31