Background Removal

A Python-based project for processing Point Cloud Data (PCD) files, featuring background removal and point cloud clustering capabilities.

Features

• Background removal from point cloud data
• Point cloud clustering (2D and 3D)
• Multiple processing methods:
  • Brute force
  • KD-tree
  • Octree
  • Spherical coordinates
  • Voxel-based
• Ground plane detection
• Visualization tools

Demo

Background Removal Result

Demonstration of the background removal process

Setup

1. Clone the repository:

   git clone https://github.com/moonmoonmoonmoon/static_background_removal.git
   cd background-removal

2. Install required packages:

   pip install -r requirements.txt

Detailed Usage Guide

Basic Usage

Run the main script with default parameters:

python main.py

Configuration Options

Key parameters that can be adjusted in the processing pipeline:

1. Background Removal:

• distance_threshold: Distance threshold for background point detection
• angle_threshold: Angle threshold for background point detection

2. Clustering:

• eps: The maximum distance between two samples for them to be considered neighbors
• min_samples: The minimum number of samples in a cluster

3. Ground Detection:

• distance_threshold: Maximum distance from point to plane for ground classification
• max_iterations: Number of RANSAC iterations

Project Structure

background-removal/
├── src/
│   ├── clustering/
│   │   ├── __init__.py
│   │   ├── cluster_2d.py
│   │   └── cluster_3d.py
│   ├── find_background/
│   │   ├── __init__.py
│   │   ├── find_cartesian.py
│   │   └── find_spherical.py
│   ├── find_ground/
│   │   ├── __init__.py
│   │   └── find_ground.py
│   ├── process_point_cloud/
│   │   ├── __init__.py
│   │   ├── process_brute.py
│   │   ├── process_kdtree.py
│   │   ├── process_octree.py
│   │   ├── process_spherical.py
│   │   ├── process_voxel.py
│   │   └── utils.py
│   ├── utils/
│   │   ├── __init__.py
│   │   └── utils.py
│   ├── visualization/
│   │   ├── __init__.py
│   │   └── visualization.py
│   └── __init__.py
├── dataset            # Contains datasets for testing
│   │   ├── static         # Static background datasets
│   │   │   └── lowell     # Lowell dataset for static scenes
│   │   ├── moving         # Moving background datasets
│   │   │   └── boston_vehicle_mount # Boston vehicle-mounted dataset for dynamic scenes
├── output             # Output folder for processed data and transmation matrix
├── LICENSE
├── README.md
├── main.py
├── requirements.txt
├── images/
└── tree_generator.py

Module Description

• clustering/: Implements 2D and 3D point cloud clustering algorithms
• find_background/: Contains algorithms for background detection in both Cartesian and spherical coordinates
• find_ground/: Implements ground plane detection algorithms and get transformation martix
• process_point_cloud/: Core processing algorithms including various optimization methods
• utils/: Common utility functions
• visualization/: Tools for visualizing point cloud data and results

Contributing Guidelines

We welcome contributions to improve the project! Here's how you can contribute:

Setting up Development Environment

1. Fork the repository
2. Create a virtual environment (using conda):

    conda create -n background-removal python=3.11
    conda activate background-removal

3. Install development dependencies:

   pip install -r requirements.txt

Code Style

• Follow PEP 8 coding standards
• Use meaningful variable and function names
• Add docstrings to functions and classes
• Keep functions focused and single-purpose

Making Contributions

1. Create a new branch for your feature:

   git checkout -b your-branch-name

2. Make your changes and commit:

   git add .
   git commit -m "Meaningful commit message"

3. Push to your fork:

   git push origin your-branch-name

4. Create a Pull Request (PR) from your fork to our main repository

Pull Request Guidelines

• Provide a clear description of the changes
• Include any relevant issue numbers
• Ensure all tests pass
• Add unit tests for new features
• Update documentation if needed

Reporting Issues

When reporting issues, please include:

• A clear description of the problem
• Steps to reproduce the issue
• Expected vs actual behavior
• System information (OS, Python version, etc.)

License

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