Potree Point Cloud Viewer Example¶

This notebook demonstrates the capabilities of the anymap PotreeMap widget for creating interactive point cloud visualizations.

What is Potree?¶

Potree is a free open-source WebGL based point cloud renderer for large point clouds. It's particularly useful for:

• LiDAR data visualization
• 3D scanning results
• Large-scale point cloud datasets
• Archaeological documentation
• Building and infrastructure modeling

Setup¶

First, import the PotreeMap widget from anymap:

In [1]:

import os
from anymap import PotreeMap

print("Potree backend loaded successfully!")

import os from anymap import PotreeMap print("Potree backend loaded successfully!")

Potree backend loaded successfully!

Basic Point Cloud Viewer¶

Create a basic Potree viewer. Note that you'll need a converted Potree point cloud dataset (with metadata.json) to load actual data:

In [2]:

# Create a basic Potree viewer
viewer = PotreeMap(
    width="100%",
    height="600px",
    background_color="#1a1a1a",  # Dark background
    point_size=1.5,
    point_size_type="adaptive",  # Adaptive point sizing
    point_shape="square",
    camera_position=[0, 0, 50],
    camera_target=[0, 0, 0],
    fov=60,
    edl_enabled=True,  # Eye Dome Lighting for better depth perception
    show_grid=True,
    grid_size=10,
    grid_color="#444444",
)

viewer

# Create a basic Potree viewer viewer = PotreeMap( width="100%", height="600px", background_color="#1a1a1a", # Dark background point_size=1.5, point_size_type="adaptive", # Adaptive point sizing point_shape="square", camera_position=[0, 0, 50], camera_target=[0, 0, 0], fov=60, edl_enabled=True, # Eye Dome Lighting for better depth perception show_grid=True, grid_size=10, grid_color="#444444", ) viewer

Out[2]:

Loading Point Cloud Data¶

To load actual point cloud data, you need a Potree-converted dataset. You can convert LAS/LAZ files using PotreeConverter:

# Example conversion command (not run in this notebook)
PotreeConverter input.las -o output_directory --output-format LAZ

For demonstration purposes, here's how you would load a point cloud:

In [3]:

# Example of loading a point cloud (requires actual converted data)
# Replace with the URL to your Potree-converted point cloud metadata.json
# viewer.load_point_cloud(
#     "https://example.com/pointclouds/your_pointcloud/metadata.json",
#     "My Point Cloud"
# )

print("To load actual point cloud data, you need a Potree-converted dataset.")
print("Point the load_point_cloud method to your metadata.json file.")

# Example of loading a point cloud (requires actual converted data) # Replace with the URL to your Potree-converted point cloud metadata.json # viewer.load_point_cloud( # "https://example.com/pointclouds/your_pointcloud/metadata.json", # "My Point Cloud" # ) print("To load actual point cloud data, you need a Potree-converted dataset.") print("Point the load_point_cloud method to your metadata.json file.")

To load actual point cloud data, you need a Potree-converted dataset.
Point the load_point_cloud method to your metadata.json file.

Point Rendering Settings¶

Customize how points are rendered in the viewer:

In [4]:

# Adjust point size
viewer.set_point_size(2.0)
print("Point size set to 2.0")

# Change point size type
viewer.set_point_size_type("fixed")  # Options: "fixed", "adaptive", "attenuation"
print("Point size type set to fixed")

# Change point shape
viewer.set_point_shape("circle")  # Options: "square", "circle"
print("Point shape set to circle")

# Adjust point size viewer.set_point_size(2.0) print("Point size set to 2.0") # Change point size type viewer.set_point_size_type("fixed") # Options: "fixed", "adaptive", "attenuation" print("Point size type set to fixed") # Change point shape viewer.set_point_shape("circle") # Options: "square", "circle" print("Point shape set to circle")

Point size set to 2.0
Point size type set to fixed
Point shape set to circle

Camera Controls¶

Control the camera position and viewing angle:

In [5]:

# Set camera position and target
viewer.set_camera_position(position=[20, 20, 30], target=[0, 0, 0])
print("Camera position updated")

# Adjust field of view
viewer.set_fov(75)
print("Field of view set to 75 degrees")

# Set clipping distances
viewer.set_clip_distances(near=0.1, far=1000)
print("Clipping distances updated")

# Set camera position and target viewer.set_camera_position(position=[20, 20, 30], target=[0, 0, 0]) print("Camera position updated") # Adjust field of view viewer.set_fov(75) print("Field of view set to 75 degrees") # Set clipping distances viewer.set_clip_distances(near=0.1, far=1000) print("Clipping distances updated")

Camera position updated
Field of view set to 75 degrees
Clipping distances updated

Visual Enhancements¶

Configure visual effects and enhancements:

In [6]:

# Configure Eye Dome Lighting (EDL) for better depth perception
viewer.enable_edl(True)
viewer.set_edl_settings(radius=1.5, strength=1.2)
print("Eye Dome Lighting configured")

# Show coordinate grid
viewer.show_coordinate_grid(show=True, size=20, color="#666666")
print("Coordinate grid enabled")

# Change background color
viewer.set_background_color("#2a2a2a")
print("Background color changed")

# Configure Eye Dome Lighting (EDL) for better depth perception viewer.enable_edl(True) viewer.set_edl_settings(radius=1.5, strength=1.2) print("Eye Dome Lighting configured") # Show coordinate grid viewer.show_coordinate_grid(show=True, size=20, color="#666666") print("Coordinate grid enabled") # Change background color viewer.set_background_color("#2a2a2a") print("Background color changed")

Eye Dome Lighting configured
Coordinate grid enabled
Background color changed

Quality Settings¶

Adjust rendering quality for performance optimization:

In [7]:

# Set rendering quality
viewer.set_quality("high")  # Options: "low", "medium", "high"
print("Rendering quality set to high")

# For large datasets, you might want to use "medium" or "low" for better performance

# Set rendering quality viewer.set_quality("high") # Options: "low", "medium", "high" print("Rendering quality set to high") # For large datasets, you might want to use "medium" or "low" for better performance

Rendering quality set to high

Multiple Point Clouds¶

Load and manage multiple point cloud datasets:

In [8]:

# Example of loading multiple point clouds
point_clouds = [
    {
        "url": "https://example.com/pointclouds/scan1/metadata.json",
        "name": "Building Scan",
    },
    {
        "url": "https://example.com/pointclouds/scan2/metadata.json",
        "name": "Terrain Scan",
    },
]

# viewer.load_multiple_point_clouds(point_clouds)
print("Multiple point clouds can be loaded simultaneously")

# Example of loading multiple point clouds point_clouds = [ { "url": "https://example.com/pointclouds/scan1/metadata.json", "name": "Building Scan", }, { "url": "https://example.com/pointclouds/scan2/metadata.json", "name": "Terrain Scan", }, ] # viewer.load_multiple_point_clouds(point_clouds) print("Multiple point clouds can be loaded simultaneously")

Multiple point clouds can be loaded simultaneously

Point Cloud Filtering¶

Filter points based on various criteria:

In [9]:

# Filter points by elevation
viewer.filter_by_elevation(min_elevation=0, max_elevation=50)
print("Elevation filter applied (0-50 units)")

# Filter by classification (LAS classification codes)
# Common classifications: 1=Unclassified, 2=Ground, 3=Low Vegetation, 4=Medium Vegetation, 5=High Vegetation, 6=Building
classifications = {
    1: True,  # Show unclassified
    2: True,  # Show ground
    3: False,  # Hide low vegetation
    4: False,  # Hide medium vegetation
    5: False,  # Hide high vegetation
    6: True,  # Show buildings
}
viewer.set_classification_visibility(classifications)
print("Classification filter applied")

# Clear all filters
# viewer.clear_filters()
# print("All filters cleared")

# Filter points by elevation viewer.filter_by_elevation(min_elevation=0, max_elevation=50) print("Elevation filter applied (0-50 units)") # Filter by classification (LAS classification codes) # Common classifications: 1=Unclassified, 2=Ground, 3=Low Vegetation, 4=Medium Vegetation, 5=High Vegetation, 6=Building classifications = { 1: True, # Show unclassified 2: True, # Show ground 3: False, # Hide low vegetation 4: False, # Hide medium vegetation 5: False, # Hide high vegetation 6: True, # Show buildings } viewer.set_classification_visibility(classifications) print("Classification filter applied") # Clear all filters # viewer.clear_filters() # print("All filters cleared")

Elevation filter applied (0-50 units)
Classification filter applied

Measurement Tools¶

Add measurement capabilities to the viewer:

In [10]:

# Add distance measurement tool
viewer.add_measurement("distance")
print("Distance measurement tool added")

# Add area measurement tool
viewer.add_measurement("area")
print("Area measurement tool added")

# Add volume measurement tool
viewer.add_measurement("volume")
print("Volume measurement tool added")

# Clear all measurements
# viewer.clear_measurements()
# print("All measurements cleared")

# Add distance measurement tool viewer.add_measurement("distance") print("Distance measurement tool added") # Add area measurement tool viewer.add_measurement("area") print("Area measurement tool added") # Add volume measurement tool viewer.add_measurement("volume") print("Volume measurement tool added") # Clear all measurements # viewer.clear_measurements() # print("All measurements cleared")

Distance measurement tool added
Area measurement tool added
Volume measurement tool added

Utility Functions¶

Demonstrate various utility functions:

In [11]:

# Fit point clouds to screen
viewer.fit_to_screen()
print("View fitted to point cloud bounds")

# Take a screenshot
viewer.take_screenshot()
print("Screenshot captured")

# Get current camera position
camera_pos = viewer.get_camera_position()
camera_target = viewer.get_camera_target()
print(f"Camera position: {camera_pos}")
print(f"Camera target: {camera_target}")

# Fit point clouds to screen viewer.fit_to_screen() print("View fitted to point cloud bounds") # Take a screenshot viewer.take_screenshot() print("Screenshot captured") # Get current camera position camera_pos = viewer.get_camera_position() camera_target = viewer.get_camera_target() print(f"Camera position: {camera_pos}") print(f"Camera target: {camera_target}")

View fitted to point cloud bounds
Screenshot captured
Camera position: [20, 20, 30]
Camera target: [0, 0, 0]

Multi-cell Rendering Test¶

Test that the viewer works correctly when displayed in multiple cells:

In [12]:

# Display the same viewer instance again
# This should maintain all the settings and state from above
viewer

# Display the same viewer instance again # This should maintain all the settings and state from above viewer

Out[12]:

In [13]:

# Change settings while displayed in multiple cells
viewer.set_background_color("#0a0a0a")
viewer.set_point_size(3.0)
viewer.set_point_shape("square")

print("Settings changed! Updates should appear on all viewer instances above.")

# Change settings while displayed in multiple cells viewer.set_background_color("#0a0a0a") viewer.set_point_size(3.0) viewer.set_point_shape("square") print("Settings changed! Updates should appear on all viewer instances above.")

Settings changed! Updates should appear on all viewer instances above.

Creating a Second Viewer Instance¶

Create a separate viewer to verify independence:

In [14]:

# Create a second, independent viewer with different settings
viewer2 = PotreeMap(
    width="100%",
    height="500px",
    background_color="#003366",  # Blue background
    point_size=1.0,
    point_size_type="attenuation",
    point_shape="circle",
    camera_position=[10, 10, 20],
    edl_enabled=False,
    show_grid=False,
)

# Load different point cloud data (if available)
# viewer2.load_point_cloud(
#     "https://example.com/pointclouds/another_scan/metadata.json",
#     "Another Point Cloud"
# )

viewer2

# Create a second, independent viewer with different settings viewer2 = PotreeMap( width="100%", height="500px", background_color="#003366", # Blue background point_size=1.0, point_size_type="attenuation", point_shape="circle", camera_position=[10, 10, 20], edl_enabled=False, show_grid=False, ) # Load different point cloud data (if available) # viewer2.load_point_cloud( # "https://example.com/pointclouds/another_scan/metadata.json", # "Another Point Cloud" # ) viewer2

Out[14]:

Cleanup¶

Clear point clouds and reset viewers:

In [15]:

# Clear all point clouds from the first viewer
viewer.clear_point_clouds()
print("Point clouds cleared from first viewer")

# Clear measurements
viewer.clear_measurements()
print("Measurements cleared")

# Clear filters
viewer.clear_filters()
print("Filters cleared")

# Clear all point clouds from the first viewer viewer.clear_point_clouds() print("Point clouds cleared from first viewer") # Clear measurements viewer.clear_measurements() print("Measurements cleared") # Clear filters viewer.clear_filters() print("Filters cleared")

Point clouds cleared from first viewer
Measurements cleared
Filters cleared

Working with Real Data¶

To work with real point cloud data in Potree:

1. Convert your data¶

Use PotreeConverter to convert LAS/LAZ files:

PotreeConverter input.las -o output_directory --output-format LAZ

2. Host the data¶

Upload the converted point cloud directory to a web server that supports CORS.

3. Load in the viewer¶

viewer.load_point_cloud(
    "https://your-server.com/pointclouds/your_data/metadata.json",
    "Your Point Cloud Name"
)

Sample Data Sources¶

You can find sample Potree-compatible datasets at:

• Potree Examples
• OpenTopography
• 3D BAG (Netherlands)

Summary¶

This notebook demonstrated the key features of the anymap PotreeMap widget:

1. Point Cloud Visualization: Interactive rendering of large point cloud datasets
2. Rendering Controls: Point size, shape, and quality settings
3. Camera Controls: Position, target, field of view, and clipping
4. Visual Effects: Eye Dome Lighting, coordinate grid, background
5. Data Management: Loading single and multiple point clouds
6. Filtering: Elevation and classification-based filtering
7. Measurements: Distance, area, and volume measurement tools
8. Multi-cell Rendering: Persistent state across notebook cells
9. Multiple Instances: Independent viewer widgets

The PotreeMap widget provides a powerful platform for interactive point cloud visualization with the full capabilities of Potree.js in a Jupyter environment.