Visualizing data¶

The DTCC platform provides an integrated viewer which has the capability to display large quantities of geometry, including meshes and point clouds.

The first example demonstrates how to visualize a point cloud colored by the x-position of the points:

import dtcc
filename = '../data/dtcc-demo-data/helsingborg-residential-2022/pointcloud.las'
pc = dtcc.io.load_pointcloud(filename)
color_data = pc.points[:,0]
pc.view(data = color_data)

The second example demonstrates how to build a city from raw data and view the building mesh with three different coloring options. The first option is the default colors calculated from vertex z-coordinates. In the second option colors are are determined based on appended array of data. In the third option the data is applied as a dictionary so that the user can flip through and visualise multiple data sets for the same mesh.

from pathlib import Path
import numpy as np
import dtcc

# Build a mesh from raw data
data_directory = Path("../data/helsingborg-residential-2022")
p = dtcc.builder.parameters.default()
p["data_directory"] = data_directory
dtcc.builder.build(p)

buildings_path = data_directory / "footprints.shp"
pointcloud_path = data_directory

origin, bounds = dtcc.builder.calculate_bounds(buildings_path, pointcloud_path, p)
city = dtcc.io.load_city(buildings_path, bounds=bounds)
pointcloud = dtcc.io.load_pointcloud(pointcloud_path, bounds=bounds)

# Build a city
city = dtcc.builder.build_city(city, pointcloud, bounds, p)

# From the city build meshes
ground_mesh, building_mesh = dtcc.builder.build_mesh(city, p)

# Toggle the viewin_option variable between 1,2,3 to try different modes of viewing.
viewing_option = 1

# View the building mesh with default colors
if viewing_option == 1:
    building_mesh.view()

# View the building mesh with data per vertex
elif viewing_option == 2:
    data = building_mesh.vertices[:, 1]

    # View the building mesh and append colors
    building_mesh.view(data=data)

# View the building mesh with a dict of colors
elif viewing_option == 3:
    # Normalise the data so it falls in the range [0,1]
    x = building_mesh.vertices[:, 0]
    y = building_mesh.vertices[:, 1]
    z = building_mesh.vertices[:, 2]

    dict_data = {"vertex_x": x, "vertex_y": y, "vertex_z": z}

    # View the building mesh and appedn colors
    building_mesh.view(data=dict_data)

The third example shows how to build a city from raw data and view its ground mesh together with the pointcloud:

from pathlib import Path
import dtcc

# Build a mesh from raw data
data_directory = Path("../data/helsingborg-residential-2022")
p = dtcc.builder.parameters.default()
p["data_directory"] = data_directory
dtcc.builder.build(p)

buildings_path = data_directory / "footprints.shp"
pointcloud_path = data_directory

origin, bounds = dtcc.builder.calculate_bounds(buildings_path, pointcloud_path, p)
city = dtcc.io.load_city(buildings_path, bounds=bounds)
pointcloud = dtcc.io.load_pointcloud(pointcloud_path, bounds=bounds)

# Build a city
city = dtcc.builder.build_city(city, pointcloud, bounds, p)

# From the city build meshes
ground_mesh, building_mesh = dtcc.builder.build_mesh(city, p)

# Remove unwanted outliers from the point cloud
pc = pointcloud.remove_global_outliers(3)

# View the building mesh togheter with the pointcloud
building_mesh.view(pc=pc)

# Alternatively the pointcloud can be viewed with the mesh as agument
# pointcloud.view(mesh=ground_mesh)

The fourth example shows how to visualise a range of meshes and/or pointclouds. A Scene object is then created to which meshes and point clouds are added. This example shows how to build a city from raw data and how to visualise the ground mesh, the building mesh and the boundary mesh together with the pointcloud that was used as input. The boundary mesh will encapsulate the entire model in visually a solid box. The clipping planes under the appearance tab in the GUI can be used to cut the boundary_mesh open to see inside. The shading mode for the boundary mesh can also be set to wireframe to make the ‘box’ see-through.

from pathlib import Path
import dtcc

# Build a mesh from raw data
data_directory = Path("../data/helsingborg-residential-2022")
p = dtcc.builder.parameters.default()
p["data_directory"] = data_directory
dtcc.builder.build(p)

buildings_path = data_directory / "footprints.shp"
pointcloud_path = data_directory

origin, bounds = dtcc.builder.calculate_bounds(buildings_path, pointcloud_path, p)
city = dtcc.io.load_city(buildings_path, bounds=bounds)
pointcloud = dtcc.io.load_pointcloud(pointcloud_path, bounds=bounds)

# Build a city
city = dtcc.builder.build_city(city, pointcloud, bounds, p)

# From the city build meshes
ground_mesh, building_mesh = dtcc.builder.build_mesh(city, p)

# Remove unwanted outliers from the point cloud
pc = pointcloud.remove_global_outliers(3)

# Create a scene and window. Add geometry to scene
scene = dtcc.viewer.Scene()
window = dtcc.viewer.Window(1200, 800)

# Add meshes with data or colors to the scene for rendering.
# To provide data for coloring the keyword argument "data="" is used.
scene.add_mesh("Building mesh", building_mesh)
scene.add_mesh("Ground mesh", ground_mesh)

# Add a pointcloud with data or colors to the scene for rendering.
# To provide data for coloring the keyword argument "data="" is used.
scene.add_pointcloud("Point cloud", pc)

# Render geometry
window.render(scene)

Viewer controls¶

Once the DTCC Viewer is running and a graphics window has appeard, the viewport can be navigated with the mouse according to:

- `Left mouse button` - Rotate the view around the camera target
- `Right mouse button` - Panning the view, thus moving the camera target
- `Scroll` - Zoom in and out at the current camera target

A GUI is also created with global controls for the whole scene under apperance which includes things like (background color and clipping planes etc). Individual GUI components are also created for each Mesh and Point Cloud that is added to the scene. To close the viewer click the regular closing symbol in the upper left corner of the window or press the ESC key.