Aller au contenu principal

Output Format Preferences

Configure how enriched LAZ files are saved to balance feature completeness with software compatibility.

Overview​

The library supports two output format strategies:

  1. Augmented LAZ (LAZ 1.4) - Full feature preservation (default)
  2. QGIS-Compatible LAZ (LAZ 1.2) - Maximum compatibility (optional)

Default Behavior​

By default, the library preserves all geometric features using LAZ 1.4 format:

# Default: Creates augmented LAZ with all features
ign-lidar-hd enrich \
  --input-dir raw_tiles/ \
  --output enriched_tiles/ \
  --mode full

Output: enriched_tiles/tile.laz (LAZ 1.4 with 30+ features)

Configuration Options​

Prefer Augmented LAZ (Default)​

Setting: PREFER_AUGMENTED_LAZ = True

  • Format: LAZ 1.4
  • Features: All 30+ geometric attributes preserved
  • Compatibility: Modern LiDAR software (CloudCompare, FME, etc.)
  • File Size: Slightly larger due to extended attributes

Benefits:

  • Complete feature set available
  • No data loss during processing
  • Future-proof format

QGIS Compatibility Mode​

Setting: AUTO_CONVERT_TO_QGIS = False (manual conversion)

For QGIS compatibility, use the conversion utility:

# Method 1: Convert after enrichment
ign-lidar-hd enrich --input-dir raw/ --output enriched/
python scripts/batch_convert_qgis.py enriched/

# Method 2: Use QGIS converter directly
python -m ign_lidar.qgis_converter enriched/tile.laz

Output:

  • enriched/tile.laz (LAZ 1.4, full features)
  • enriched/tile_qgis.laz (LAZ 1.2, QGIS-compatible)

Format Comparison​

FeatureAugmented LAZQGIS LAZ
Format VersionLAZ 1.4LAZ 1.2
Geometric Features30+ attributesSelected attributes
File Size~310MB~305MB
QGIS CompatibilityModern versionsAll versions
CloudCompare✅ Full support✅ Full support
FME✅ Full support✅ Full support
Custom Analysis✅ All features⚠️ Limited features

Available Features by Format​

Augmented LAZ (Full Feature Set)​

Geometric Features:

  • normal_x, normal_y, normal_z - Surface normals
  • curvature - Surface curvature
  • planarity, linearity, sphericity - Shape descriptors
  • roughness - Surface texture
  • anisotropy - Directional variation

Building-Specific Features:

  • verticality - Wall orientation score
  • wall_score, roof_score - Component probabilities
  • height_above_ground - Normalized elevation
  • density - Local point density
  • vertical_std - Height variation

Advanced Features:

  • eigenvalue_1, eigenvalue_2, eigenvalue_3 - Principal components
  • num_points_2m - Neighborhood size
  • neighborhood_extent - Spatial extent
  • height_extent_ratio - Geometric ratios

QGIS LAZ (Compatible Subset)​

Core Features (guaranteed compatibility):

  • normal_x, normal_y, normal_z
  • curvature
  • planarity, linearity, sphericity
  • height_above_ground
  • wall_score, roof_score
  • verticality

Excluded Features (compatibility limitations):

  • Advanced eigenvalue decomposition
  • Complex neighborhood statistics
  • Extended geometric ratios

When to Use Each Format​

Use Augmented LAZ When:​

  • Research applications requiring full feature set
  • Custom analysis with machine learning models
  • Advanced visualization in CloudCompare or specialized tools
  • Future processing where feature completeness matters
  • Archival storage for long-term data preservation

Use QGIS LAZ When:​

  • GIS workflows primarily using QGIS
  • Visualization focused on basic geometric properties
  • Sharing data with users who have older software
  • Web mapping applications with format constraints
  • Teaching/demos where compatibility is critical

Configuration Examples​

Research/Development Workflow​

# Full feature extraction for ML research
ign-lidar-hd enrich \
  --input-dir raw_tiles/ \
  --output research_data/ \
  --mode full

# Keep all 30+ features in LAZ 1.4 format
# Use for: scikit-learn, PyTorch, custom analysis

GIS/Visualization Workflow​

# Basic enrichment
ign-lidar-hd enrich \
  --input-dir raw_tiles/ \
  --output enriched_tiles/ \
  --mode full

# Convert for QGIS
python scripts/batch_convert_qgis.py enriched_tiles/

# Result: Both formats available
ls enriched_tiles/
# tile.laz          (full features)
# tile_qgis.laz     (QGIS-compatible)

Production Workflow​

# Process with smart skip detection
ign-lidar-hd enrich \
  --input-dir large_dataset/ \
  --output processed/ \
  --mode full

# Selectively convert tiles needing QGIS compatibility
python -m ign_lidar.qgis_converter processed/priority_tiles/

Programmatic Configuration​

Python API​

from ign_lidar import LiDARProcessor, config

# Option 1: Use configuration
config.PREFER_AUGMENTED_LAZ = True
config.AUTO_CONVERT_TO_QGIS = False

processor = LiDARProcessor()
processor.enrich_file('input.laz', 'output/')

# Option 2: Direct format control
processor = LiDARProcessor(
    output_format='augmented'  # or 'qgis'
)

Configuration File​

Create config/local_settings.py:

# Format preferences
PREFER_AUGMENTED_LAZ = True
AUTO_CONVERT_TO_QGIS = False

# Processing preferences
DEFAULT_MODE = 'full'
DEFAULT_WORKERS = 4

# Output preferences
PRESERVE_DIRECTORY_STRUCTURE = True
COPY_METADATA_FILES = True

File Naming Conventions​

Standard Output​

enriched_tiles/
├── LIDARHD_FXX_0123_4567_LA93_IGN69_2020.laz    # Augmented LAZ
└── metadata/
    └── LIDARHD_FXX_0123_4567_LA93_IGN69_2020.json

With QGIS Conversion​

enriched_tiles/
├── LIDARHD_FXX_0123_4567_LA93_IGN69_2020.laz         # Augmented LAZ
├── LIDARHD_FXX_0123_4567_LA93_IGN69_2020_qgis.laz    # QGIS-compatible
└── metadata/
    └── LIDARHD_FXX_0123_4567_LA93_IGN69_2020.json

Quality Verification​

Check Format Version​

import laspy

# Check file format
las = laspy.read('enriched_tile.laz')
print(f"LAZ version: {las.header.version}")
print(f"Point format: {las.point_format.id}")
print(f"Extra dimensions: {las.point_format.extra_dimension_names}")

Validate QGIS Compatibility​

# Test QGIS compatibility
python scripts/validation/test_qgis_compatibility.py enriched_tile_qgis.laz

# Expected output:
# ✅ LAZ 1.2 format detected
# ✅ Compatible point format (ID 6)
# ✅ Standard dimensions present
# ✅ Compatible extra dimensions: 8 found
# ✅ File should load properly in QGIS

Troubleshooting​

QGIS Won't Load File​

Issue: "Unsupported file format" in QGIS

Solution: Convert to QGIS format

python -m ign_lidar.qgis_converter problematic_file.laz
# Creates: problematic_file_qgis.laz

Missing Features in Analysis​

Issue: Can't find expected attributes

Cause: Using QGIS-compatible format instead of full format

Solution: Use augmented LAZ for analysis

# Load full feature set
las = laspy.read('tile.laz')  # Not tile_qgis.laz
features = las.point_format.extra_dimension_names
print(f"Available features: {len(features)}")

Large File Sizes​

Issue: Files larger than expected

Cause: Full feature set increases file size

Solution:

  1. Use QGIS format for storage-constrained applications
  2. Compress with higher LAZ compression levels
  3. Filter to essential features only

Performance Considerations​

Processing Speed​

  • Augmented LAZ: Faster (no conversion overhead)
  • QGIS LAZ: Slower (requires attribute filtering/conversion)

Storage Requirements​

  • Augmented LAZ: ~5-10% larger files
  • QGIS LAZ: Standard LAZ file sizes
  • Both formats: 2x storage if keeping both

Memory Usage​

  • Augmented LAZ: Higher memory during processing
  • QGIS LAZ: Lower memory (fewer attributes)

See Also​