Person Re-identification Using Clustering

This repository contains Jupyter Notebooks and resources for person re-identification using clustering techniques. The process involves the following steps:

1. Converting video to frames.
2. Detecting bounding boxes around persons in the frames.
3. Cropping the detected objects (persons) from the frames.
4. Clustering the cropped objects using k-means clustering.

Repository Structure

• video_to_frames1.ipynb: Notebook to convert video files into individual frames.
• frames_to_bounding_boxes.ipynb: Notebook to detect bounding boxes in frames.
• bounding_boxes_to_cropped_objects.ipynb: Notebook to crop objects from frames based on bounding boxes.
• K-Means_clustering.ipynb: Notebook to perform k-means clustering on cropped objects.
• data: Directory containing example data for testing the notebooks.
• requirements.txt: List of dependencies required to run the notebooks.

Installation

First, clone the repository and navigate to the project directory:

git clone https://github.com/ankithagithub/Person_Re-Identification_using_clustering.git
cd Person_Re-Identification_using_clustering
pip install -r requirements.txt

Usage

• Step 1: Convert Video to Frames

  Open video_to_frames1.ipynb in Jupyter Notebook or Jupyter Lab and run the cells to convert a video file into individual frames.

• Step 2: Detect Bounding Boxes

  Open frames_to_bounding_boxes.ipynb in Jupyter Notebook or Jupyter Lab and run the cells to detect bounding boxes in the frames.

• Step 3: Crop Objects from Frames

  Open bounding_boxes_to_cropped_objects.ipynb in Jupyter Notebook or Jupyter Lab and run the cells to crop objects from frames based on the detected bounding boxes.

• Step 4: Perform Clustering

  Open K-Means_clustering.ipynb in Jupyter Notebook or Jupyter Lab and run the cells to perform k-means clustering on the cropped objects.

License

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

Contributing

Contributions are welcome! Please open an issue or submit a pull request if you have any suggestions or improvements.

Acknowledgments

Special thanks to all the contributors and the open-source community for providing the tools and libraries used in this project.