Skip to main content

Processor API

Complete API reference for the IGN LiDAR HD processor module.

Processor Class​

The main processor class for handling LiDAR data processing workflows.

Constructor​

from ign_lidar import Processor

processor = Processor(config=None, verbose=True)

Parameters:

  • config (Config, optional): Configuration object with processing parameters
  • verbose (bool): Enable verbose logging output

Methods​

process_tile()​

Process a single LAS/LAZ tile with feature extraction.

def process_tile(self, input_path: str, output_path: str = None) -> dict:

Parameters:

  • input_path (str): Path to input LAS/LAZ file
  • output_path (str, optional): Path for output file

Returns:

  • dict: Processing results with metrics and status

Example:

result = processor.process_tile("input.las", "output_enriched.las")
print(f"Processed {result['points_count']} points")

batch_process()​

Process multiple tiles in batch mode.

def batch_process(self, tile_list: list, output_dir: str) -> list:

Parameters:

  • tile_list (list): List of input file paths
  • output_dir (str): Output directory for processed files

Returns:

  • list: List of processing results for each tile

extract_features()​

Extract geometric and radiometric features from point cloud.

def extract_features(self, points: numpy.ndarray,
                    feature_types: list = None) -> numpy.ndarray:

Parameters:

  • points (numpy.ndarray): Point cloud data
  • feature_types (list, optional): List of features to extract

Returns:

  • numpy.ndarray: Array with extracted features

Available feature types:

  • height_above_ground: Normalized height
  • intensity_normalized: Normalized intensity
  • local_density: Local point density
  • surface_roughness: Surface roughness measure
  • planarity: Planarity coefficient
  • sphericity: Sphericity coefficient

classify_points()​

Classify points into semantic categories.

def classify_points(self, points: numpy.ndarray,
                   features: numpy.ndarray) -> numpy.ndarray:

Parameters:

  • points (numpy.ndarray): Point cloud data
  • features (numpy.ndarray): Extracted features

Returns:

  • numpy.ndarray: Classification labels

Configuration Options​

Processing Parameters​

from ign_lidar import Config

config = Config(
    chunk_size=1000000,
    overlap_size=50,
    classification_model="random_forest",
    feature_radius=2.0,
    min_points_per_class=100
)

Performance Tuning​

config = Config(
    use_gpu=True,
    max_memory_gb=16,
    num_workers=8,
    cache_features=True
)

Error Handling​

The processor handles common errors gracefully:

try:
    result = processor.process_tile("input.las")
except FileNotFoundError:
    print("Input file not found")
except MemoryError:
    print("Insufficient memory for processing")
except ValueError as e:
    print(f"Invalid parameter: {e}")

Performance Tips​

  1. Batch Processing: Process multiple tiles together for better efficiency
  2. Memory Management: Set appropriate chunk sizes for available RAM
  3. GPU Acceleration: Enable GPU processing for large datasets
  4. Feature Caching: Cache computed features for repeated use