Skip to main content

CLI Enrich Command Reference

The ign-lidar enrich command adds building component features to LiDAR point clouds, enhancing them with geometric and architectural information.

Syntax​

ign-lidar enrich [OPTIONS] INPUT_PATH OUTPUT_PATH

Basic Usage​

Enrich a Single File​

ign-lidar enrich input.las enriched_output.las

Enrich with RGB Data​

ign-lidar enrich --rgb-path orthophoto.tif input.las output.las

Batch Processing​

ign-lidar enrich --batch tiles/*.las --output-dir enriched/

Command Options​

Input/Output Options​

INPUT_PATH (required)​

Path to input LAS/LAZ file or directory for batch processing.

# Single file
ign-lidar enrich tile_001.las output.las

# Directory (batch mode)
ign-lidar enrich tiles/ --output-dir enriched/

OUTPUT_PATH (required for single files)​

Output path for enriched point cloud.

--output-dir, -o (batch mode)​

Output directory for batch processing.

ign-lidar enrich tiles/ --output-dir /path/to/enriched/

--output-format​

Output file format.

Options: las, laz, ply, h5 Default: las

ign-lidar enrich input.las output.h5 --output-format h5

Processing Options​

--features, -f​

Specify which features to extract.

Options:

  • geometric: Basic geometric features (planarity, linearity, etc.)
  • architectural: Building-specific features (walls, roofs, etc.)
  • all: All available features
  • custom: User-defined feature set
# Extract only geometric features
ign-lidar enrich --features geometric input.las output.las

# Extract all features
ign-lidar enrich --features all input.las output.las

# Custom feature set
ign-lidar enrich --features planarity,height,normal_z input.las output.las

--neighborhood-size, -n​

Neighborhood radius for feature computation (meters).

Default: 1.0

ign-lidar enrich --neighborhood-size 2.0 input.las output.las

--min-building-points​

Minimum points required to classify as building.

Default: 50

ign-lidar enrich --min-building-points 100 input.las output.las

RGB Integration Options​

--rgb-path, -r​

Path to orthophoto for color augmentation.

ign-lidar enrich --rgb-path orthophoto.tif input.las output.las

--rgb-interpolation​

Interpolation method for RGB assignment.

Options: nearest, bilinear, bicubic Default: bilinear

ign-lidar enrich --rgb-path ortho.tif --rgb-interpolation bicubic input.las output.las

--rgb-bands​

Specify which bands to extract from orthophoto.

Default: 1,2,3 (RGB)

# Include infrared band
ign-lidar enrich --rgb-path ortho.tif --rgb-bands 1,2,3,4 input.las output.las

Performance Options​

--gpu, -g​

Enable GPU acceleration.

ign-lidar enrich --gpu input.las output.las

--gpu-memory-fraction​

Fraction of GPU memory to use.

Default: 0.7

ign-lidar enrich --gpu --gpu-memory-fraction 0.9 input.las output.las

--batch-size, -b​

Processing batch size for memory management.

Default: 100000

ign-lidar enrich --batch-size 50000 input.las output.las

--num-workers, -w​

Number of parallel workers for processing.

Default: Number of CPU cores

ign-lidar enrich --num-workers 8 input.las output.las

Quality Control Options​

--validate​

Perform validation checks on output.

ign-lidar enrich --validate input.las output.las

--quality-report​

Generate quality assessment report.

ign-lidar enrich --quality-report report.json input.las output.las

--preserve-classification​

Keep original point classifications.

ign-lidar enrich --preserve-classification input.las output.las

Architectural Analysis Options​

--architectural-style​

Specify architectural style for enhanced analysis.

Options: haussmanian, traditional, contemporary, industrial

ign-lidar enrich --architectural-style haussmanian input.las output.las

--region​

Geographic region for style adaptation.

Options: ile_de_france, provence, brittany, alsace

ign-lidar enrich --region ile_de_france input.las output.las

--building-type​

Expected building types in dataset.

Options: residential, commercial, industrial, mixed

ign-lidar enrich --building-type residential input.las output.las

Configuration File​

Use a YAML configuration file for complex processing scenarios.

Configuration Example​

# enrich_config.yaml
processing:
  features: ["geometric", "architectural"]
  neighborhood_size: 1.5
  min_building_points: 75

rgb:
  enabled: true
  interpolation: "bilinear"
  bands: [1, 2, 3]

performance:
  use_gpu: true
  gpu_memory_fraction: 0.8
  batch_size: 75000
  num_workers: 6

architectural:
  style: "haussmanian"
  region: "ile_de_france"
  building_type: "residential"

quality:
  validate: true
  generate_report: true

Using Configuration File​

ign-lidar enrich --config enrich_config.yaml input.las output.las

Advanced Examples​

High-Quality Building Analysis​

ign-lidar enrich \
  --features all \
  --rgb-path orthophoto.tif \
  --architectural-style haussmanian \
  --region ile_de_france \
  --neighborhood-size 1.5 \
  --gpu \
  --validate \
  --quality-report quality.json \
  input.las output.las

Batch Processing with GPU​

ign-lidar enrich \
  --batch tiles/*.las \
  --output-dir enriched/ \
  --features geometric,architectural \
  --gpu \
  --batch-size 200000 \
  --num-workers 4

Memory-Optimized Processing​

ign-lidar enrich \
  --batch-size 25000 \
  --gpu-memory-fraction 0.5 \
  --num-workers 2 \
  input_large.las output.las

Output Information​

Added Point Fields​

The enrich command adds the following fields to point clouds:

Field NameTypeDescription
planarityfloatSurface planarity (0-1)
linearityfloatLinear structure strength (0-1)
sphericityfloat3D compactness (0-1)
height_above_groundfloatNormalized height (meters)
building_componentuint8Component class (0=ground, 1=wall, 2=roof)
architectural_styleuint8Detected architectural style
normal_x, normal_y, normal_zfloatSurface normal vectors

RGB Fields (when enabled)​

Field NameTypeDescription
red, green, blueuint16RGB color values
infrareduint16Near-infrared (if available)
material_classuint8Material classification

Error Handling​

Common Error Messages​

"GPU memory insufficient"​

Solution: Reduce batch size or GPU memory fraction

ign-lidar enrich --gpu-memory-fraction 0.5 --batch-size 50000 input.las output.las

"RGB file not found"​

Solution: Check orthophoto path and file permissions

ls -la /path/to/orthophoto.tif

"Insufficient points for feature extraction"​

Solution: Lower minimum building points threshold

ign-lidar enrich --min-building-points 25 input.las output.las

Debugging Options​

--verbose, -v​

Enable detailed logging output.

ign-lidar enrich --verbose input.las output.las

--debug​

Enable debug mode with extensive logging.

ign-lidar enrich --debug input.las output.las

--log-file​

Save logs to file.

ign-lidar enrich --log-file enrich.log input.las output.las

Performance Benchmarks​

Processing Times (approximate)​

PointsFeaturesGPUCPUSpeedup
1MGeometric30s4min8x
1MAll + RGB45s8min11x
10MGeometric3min35min12x
10MAll + RGB5min75min15x

Memory Usage​

  • CPU Mode: ~6GB RAM per 10M points
  • GPU Mode: ~3GB GPU + 2GB RAM per 10M points
  • Batch Processing: Configurable memory footprint

See Also​