Bathy Subsample

A Python package for processing bathymetric point clouds using isolation forests and voxel grids. This tool helps in subsampling bathymetric data while preserving important features and adhering to IHO standards.

Features

• Point cloud anomaly detection using isolation forests
• Adaptive voxel grid processing
• Multi-modal depth analysis
• TVU-based uncertainty handling
• Parallel processing support
• Comprehensive visualization tools
• IHO standards compliance

Installation

# Clone the repository
git clone https://github.com/avfonseca/Bathy_Sumbsample.git
cd Bathy_Subsample

# Install in development mode
pip install -e .

Requirements

• Python 3.6+
• NumPy
• SciPy
• scikit-learn
• matplotlib
• joblib
• tqdm
• seaborn
• eif -> https://github.com/sahandha/eif

Usage

Basic usage example:

from bathy_subsample import IsolationGrid

# Create processor instance
processor = IsolationGrid(
    group_size=1000,
    voxel_x_size=1.0,
    voxel_y_size=1.0,
    mode_probability_threshold=0.6,
    anomaly_threshold=0.5,
    min_points_for_mode=5,
    max_modes=2,
    save_intermediate_files=True
)

# Process point cloud
final_points, stats = processor.process('input_points.xyz')

Configuration Parameters

• group_size: Number of points to process in each leaf (default: 1000)
• voxel_x_size: Size of voxels in X dimension in meters (default: 1.0)
• voxel_y_size: Size of voxels in Y dimension in meters (default: 1.0)
• anomaly_threshold: Threshold for anomaly scores (default: 0.5)
• mode_probability_threshold: Minimum probability to assign point to a mode
• min_points_for_mode: Minimum points needed for mode fitting
• max_modes: Maximum number of modes to fit (default: 1)

Output

The processor generates:

• Processed point cloud file
• Visualization plots
• Statistical analysis
• Processing reports

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Citation

If you use this software in your research, please cite:

@software{bathy_subsample,
  author = {Adriano Fonseca},
  title = {Bathy Subsample: A Tool for Bathymetric Point Cloud Processing},
  year = {2024},
  url = {https://github.com/avfonseca/bathy_subsample}
}