lidarlens 0.1.4

A Python package for 3D LiDAR point cloud processing, analysis, terrain modelling, and visualization.

LidarLens

LidarLens is a powerful Python package for processing, analyzing, and visualizing 3D LiDAR point cloud data. It provides a clean, pythonic API for common tasks such as I/O, noise removal, ground extraction, DTM/DSM generation, and terrain analysis.

Features

• I/O: Read and write LAS/LAZ files (wraps laspy and open3d).
• Preprocessing: Voxel downsampling, statistical/radius outlier removal.
• Segmentation:
  • Ground extraction using CSF (Cloth Simulation Filter) or RANSAC.
  • Building and vegetation segmentation (heuristic).
  • DBSCAN clustering.
• Terrain Modelling:
  • Generate DTM (Digital Terrain Model) and DSM (Digital Surface Model).
  • Compute Canopy Height Models (CHM).
  • Generate vector contours.
• Analysis:
  • Point density maps, roughness, slope, aspect, and hillshade.
  • Cloud-to-cloud distance comparison.
  • Volume computation and cross-section profiling.
• Geometry:
  • Normal estimation and orientation.
  • Mesh surface reconstruction (Poisson, Ball Pivoting).
  • ICP Registration.
• Visualization: Helper functions for plotting point clouds, DEMs, and histograms.
• CLI: Comprehensive command-line interface for batch processing.

Installation

pip install lidarlens

To install with CLI support:

pip install lidarlens[cli]

To install with Deep Learning support (coming soon):

pip install lidarlens[dl]

Quick Start

Python API: DEM Generation Pipeline

This example demonstrates how to import a raw point cloud, remove noise, classify ground points, and generate a Digital Terrain Model (DTM).

import lidarlens as ll

# 1. Import Data
# Load your point cloud file (supports .las and .laz)
# Returns the point cloud, header, VLRs, and coordinate reference system (WKT)
pcd, header, vlrs, wkt = ll.read("raw_scan.laz")
print(f"Loaded {header.point_count} points")

# 2. Data Processing
# Remove statistical outliers (noise removal)
pcd_clean = ll.denoise(pcd, method="statistical", nb_neighbors=20, std_ratio=2.0)

# Downsample to 0.5m voxels for faster processing
pcd_down = ll.downsample(pcd_clean, voxel_size=0.5)

# 3. Ground Extraction
# Separate likely ground points from non-ground objects
ground_pcd, non_ground_pcd, plane_model = ll.extract_ground(pcd_down)

# 4. Generate DEM (Digital Terrain Model)
# Rasterize ground points into a 1.0m resolution grid
dtm, meta = ll.generate_dtm(ground_pcd, resolution=1.0)

# 5. Visualize and Save
# Save as a colorized PNG for quick visualization
ll.dem_to_png(dtm, meta, "output_dtm.png", colormap="terrain")

# Export to GeoTIFF for use in GIS software (requires rasterio)
# We pass the 'wkt' to ensure the output is properly georeferenced
ll.dem_to_geotiff(dtm, meta, "output_dtm.tif")

Command Line Interface

# Info
lidarlens info input.laz

# Processing pipeline
lidarlens downsample input.laz -v 0.5 -o downsampled.laz
lidarlens denoise downsampled.laz -o clean.laz
lidarlens ground clean.laz -o ground.laz

# Terrain products
lidarlens dtm ground.laz -o dtm.png
lidarlens contours ground.laz -i 2.0 -o contours.geojson