🏛️ VoxCity Landmark Visibility Analysis

Analyze visibility of specific landmarks (buildings) from all locations in your study area using 3D ray-casting.

Use Cases

  • Wayfinding research - Which landmarks are visible from different locations?

  • Urban planning - Assess visual impact of new buildings

  • Real estate - Evaluate views from different properties

  • Tourism - Map visibility of attractions

Landmark Selection Methods

Method

When to Use

landmark_building_ids

You know specific building IDs

landmark_polygon

Define custom area containing landmarks

Auto-select (center)

Let VoxCity pick buildings near center

Prerequisites

pip install voxcity

# %pip install voxcity

from voxcity.generator import get_voxcity
from voxcity.geoprocessor.draw import draw_rectangle_map_cityname
from voxcity.simulator.view import get_landmark_visibility_map

meshsize = 5
cityname = "Tokyo, Japan"

rectangle_vertices = [
    (139.760, 35.680),  # SW
    (139.760, 35.690),  # NW
    (139.770, 35.690),  # NE
    (139.770, 35.680)   # SE
]

city = get_voxcity(
    rectangle_vertices,
    meshsize=meshsize,
    building_source='OpenStreetMap',
    land_cover_source='OpenStreetMap',
    canopy_height_source='High Resolution 1m Global Canopy Height Maps',
    dem_source='DeltaDTM',
    output_dir='output/landmark_demo'
)

# Access building_gdf from the VoxCity object
building_gdf = city.extras.get('building_gdf', None)
len(building_gdf) if building_gdf is not None else 0

🎯 Select Landmarks

Three ways to specify which buildings are “landmarks”:

  1. By IDs: Provide landmark_building_ids - list of building indices from GeoDataFrame

  2. By Polygon: Provide landmark_polygon - GeoJSON-style polygon, all buildings inside become landmarks

  3. Auto-center: Omit both parameters - function picks buildings at rectangle center

landmark_kwargs = {
    "view_point_height": 1.5,
    "colormap": "cool",
    "obj_export": True,
    "output_directory": "output/landmark_demo",
    "output_file_name": "landmark_visibility",
    "alpha": 1.0,
}

# Example 1: Pick by IDs (take first few as sample)
ids_sample = building_gdf.head(3).index.tolist() if building_gdf is not None else []

landmark_vis_map, voxcity_grid_marked = get_landmark_visibility_map(
    city,
    building_gdf=building_gdf,
    landmark_building_ids=ids_sample,
    **landmark_kwargs
)

landmark_vis_map.shape

# Example 2: Auto-select using rectangle center (omit IDs and polygon)
landmark_vis_map_auto, _ = get_landmark_visibility_map(
    city,
    building_gdf=building_gdf,
    rectangle_vertices=rectangle_vertices,
    **landmark_kwargs
)

landmark_vis_map_auto.shape

📊 Interpreting Results

  • Visibility values: 0 = not visible, 1 = fully visible

  • Output: Colored OBJ mesh showing visibility from each location

  • voxcity_grid_marked: 3D voxel grid with landmarks highlighted

Next Steps

  • demo_view.ipynb - Green View Index & Sky View Index

  • demo_solar.ipynb - Solar irradiance simulation

  • demo_3d_visualization.ipynb - Advanced 3D visualization