voxel-obstacle-detection 0.1.0

A Python package for obstacle detection in voxel maps using Open3D and raycasting

Obstacle Detection

Obstacle Detection is a Python package designed to facilitate obstacle detection in voxelized environments using raycasting techniques. It leverages Open3D for 3D geometry processing and is optimized for robotics and visualization tasks.

---

Features

• Voxelized Mesh Handling: Import and process 3D voxel maps for obstacle detection.
• Raycasting: Cast rays in customizable directions to detect obstacles.
• Visualization: Render voxel maps, raycasting results, and directional vectors using Open3D.
• Dominant Direction Analysis: Analyze the dominant direction in a voxel map for robot orientation.
• Flexible APIs: Integrate with other robotics or visualization frameworks.

---

Installation

Install the package via pip:

pip install obstacle-detection

## Dependencies

The package relies on the following dependencies:

- `numpy`
- `open3d`
- `scikit-learn`
- `urdfpy`

All required dependencies will be installed automatically when using pip install.

---

Usage

Example 1: Visualizing a Voxel Map

from voxel_object_detection.Obstacle_Detection import Obstacle_Detection

od = Obstacle_Detection()

od.show_voxel_map(file_path="./examples/data/voxel_map_example.json")

Example 2: Detect Obstacles in a Given Direction

from voxel_object_detection.Obstacle_Detection import Obstacle_Detection

max_distance = 30
num_rays = 100
angle_range = 30
threshold = 0.1
file_path = "./examples/data/voxel_map_example.json"
move_direction = [1, 0, 0]

od = Obstacle_Detection(
    max_distance=max_distance,
    num_rays=num_rays,
    angle_range=angle_range
)

voxel_map = od.load_voxel_map_from_json(file_path)

scene, dominant_direction = od.preprocess_voxel_map(voxel_map, render=False)

path_is_clear = od.is_path_clear(
    dominant_direction=dominant_direction,
    relative_walking_vector=move_direction,
    scene=scene,
    threshold=threshold,
)

if path_is_clear:
    print("Path is clear!")
else:
    print("Obstacle detected!")

API Reference

ObstacleDetection

Methods

• __init__(voxel_map_path: str)

  • Initialize the class with a JSON voxel map file.

• visualize_voxel_map()

  • Render the voxel map and raycasting results using Open3D.

• check_clear_path(center: list, dominant_direction: list, relative_direction: list, max_distance: float) -> bool

  • Check if there is a clear path in a given direction, relative to the dominant direction.

---

Contributing

Contributions are welcome! Please fork the repository and submit a pull request for any features, fixes, or improvements.

---

License

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

---

Acknowledgments

• Open3D: For powerful 3D visualization and raycasting tools.
• NumPy: For numerical computations.
• Scikit-learn: For PCA and other analysis utilities.