gaussian-reconstruction 1.2.1

3D gaussian splatting recosntruction

Gaussian Reconstruction

Gaussian Reconstruction is a Python package for 3D Gaussian splatting reconstruction from video inputs, designed for efficient and high-quality 3D model generation.

Installation and Setup

Follow these steps to set up the environment and install the package.

Step 1: Create the Conda Environment

First, create the gaussian_splatting environment using the provided environment.yml file:

conda env create -f gsrecon/environment.yml

Step 2: Activate the Environment

Activate the newly created environment:

conda activate gaussian_splatting

Step 3: Install Gaussian Reconstruction Package

Install the gaussian-reconstruction package using pip:

pip install gaussian-reconstruction

Additional Dependencies

To ensure successful operation, the following dependencies are required (these are included in the package requirements, but you may install them manually if needed):

• OpenCV (cv2): pip install opencv-python
• PyColmap: pip install pycolmap

Usage

Once the package is installed, you can start the conversion process.

Step 4: Import the Library

Import the VideoToPLYConverter class from the package:

from gsrecon.converter import VideoToPLYConverter

Step 5: Create a Converter Instance

Create an instance of VideoToPLYConverter with the desired method (e.g., pycolmap):

converter = VideoToPLYConverter(method='pycolmap')

Step 6: Start the Conversion

Use the convert_video_to_ply function to start the conversion process. Provide the path to the video file and set the extraction rate (frames per second):

video_path = "path/to/your/video.mp4"  # Replace with your video file path
extraction_rate = None  # Set the frame extraction rate; defaults to None (extract all frames unless specified, e.g., 1 for 1 FPS)
converter.convert_video_to_ply(video_path, extraction_rate)

License

This project is licensed under the terms specified in the LICENSE file.

---

This documentation provides a step-by-step guide to installing, setting up, and using the Gaussian Reconstruction package for 3D Gaussian splatting reconstruction.