Add script to visualize tiles.

README.md

@@ -89,6 +89,10 @@ Then launch Python shell:
     # These group together many 1.25 m/pixel 512x512 tiles into one tar file.
     print(f"{epsg_code}_{col//512//32}_{row//512//32}")
 
+So then you would download the tar file from the second prefix, extract it, and look at the file with name matching the first prefix.
+
+See visualize_tile.py for example of visualizing the data at a particular tile.
+
 
 Sentinel-2
 ----------

visualize_tile.py

@@ -0,0 +1,95 @@
+"""
+This script visualizes a tile.
+Example usage:
+python visualize_tile.py 32610_859_-8247 /path/to/dataset/ /path/to/output/
+"""
+
+import json
+import os
+import sys
+
+import numpy as np
+from PIL import Image
+import rasterio
+import rasterio.features
+
+label = sys.argv[1]
+root_dir = sys.argv[2]
+out_dir = sys.argv[3]
+
+# Landsat.
+with rasterio.open(os.path.join(root_dir, "landsat", f"{label}_8.tif")) as src:
+    array = src.read()
+for idx in range(16):
+    img = np.clip((array[idx, :, :].astype(np.float64) - 5000) / 20, 0, 255).astype(np.uint8)
+    img = img.repeat(axis=0, repeats=8).repeat(axis=1, repeats=8)
+    Image.fromarray(img).save(os.path.join(out_dir, f"{label}_landsat{idx}.png"))
+
+# NAIP.
+array = np.array(Image.open(os.path.join(root_dir, "naip", f"{label}.png")))
+Image.fromarray(array[:, :, 0:3]).save(os.path.join(out_dir, f"{label}_naip.png"))
+
+# Old NAIP.
+array = np.array(Image.open(os.path.join(root_dir, "oldnaip", f"{label}.png")))
+Image.fromarray(array[:, :, 0:3]).save(os.path.join(out_dir, f"{label}_oldnaip.png"))
+
+# OpenStreetMap.
+with open(os.path.join(root_dir, "openstreetmap", f"{label}.geojson")) as f:
+    data = json.load(f)
+category_colors = {
+    "river": [0, 0, 255],
+    "road": [255, 255, 255],
+    "building": [255, 255, 0],
+    "parking": [255, 0, 0],
+    "leisure_park": [144, 238, 144],
+    "solar": [128, 128, 128],
+}
+category_selectors = {
+    "leisure_park": lambda feat: feat["properties"]["category"] == "leisure" and feat["properties"].get("leisure") == "park",
+    "solar": lambda feat: feat["properties"]["category"] == "power_plant" and feat["properties"].get("plant:source") == "solar",
+}
+img = np.zeros((512, 512, 3), dtype=np.uint8)
+for category, color in category_colors.items():
+    selector = category_selectors.get(category, lambda feat: feat["properties"]["category"] == category)
+    geometries = [feat["geometry"] for feat in data["features"] if selector(feat)]
+    if len(geometries) == 0:
+        continue
+    mask = rasterio.features.rasterize(geometries, out_shape=(512, 512))
+    img[mask > 0] = color
+Image.fromarray(img).save(os.path.join(out_dir, f"{label}_openstreetmap.png"))
+
+# Sentinel-1.
+with rasterio.open(os.path.join(root_dir, "sentinel1", f"{label}.tif")) as src:
+    array = src.read()
+img = np.clip((array[0, :, :] + 20) * 10, 0, 255).astype(np.uint8)
+img = img.repeat(axis=0, repeats=8).repeat(axis=1, repeats=8)
+Image.fromarray(img).save(os.path.join(out_dir, f"{label}_sentinel1.png"))
+
+# Sentinel-2.
+with rasterio.open(os.path.join(root_dir, "sentinel2", f"{label}_8.tif")) as src:
+    array = src.read()
+for idx in range(8):
+    img = np.clip(array[(idx*4+2, idx*4+1, idx*4+0), :, :].transpose(1, 2, 0) / 10, 0, 255).astype(np.uint8)
+    img = img.repeat(axis=0, repeats=8).repeat(axis=1, repeats=8)
+    Image.fromarray(img).save(os.path.join(out_dir, f"{label}_sentinel2_{idx}.png"))
+
+# WorldCover.
+array = np.array(Image.open(os.path.join(root_dir, "worldcover", f"{label}.png")))
+img = np.zeros((512, 512, 3), dtype=np.uint8)
+category_colors = {
+    10: [0, 100, 0],
+    20: [255, 187, 34],
+    30: [255, 255, 76],
+    40: [240, 150, 255],
+    50: [250, 0, 0],
+    60: [180, 180, 180],
+    70: [240, 240, 240],
+    80: [0, 100, 200],
+    90: [0, 150, 160],
+    95: [0, 207, 117],
+    100: [250, 230, 160],
+}
+for category, color in category_colors.items():
+    mask = (array == category).repeat(axis=0, repeats=8).repeat(axis=1, repeats=8)
+    img[mask] = color
+Image.fromarray(img).save(os.path.join(out_dir, f"{label}_worldcover.png"))