Hemato-Vision

A web-based application for automated blood cell classification, designed to predict the type of blood cells (eosinophil, lymphocyte, monocyte, neutrophil) using a pre-trained deep learning model.

Overview

Hemato-Vision is a machine learning application developed to assist in hematological analysis by classifying blood cell images. It uses a TensorFlow deep learning model (Blood_Cell.h5) trained on labeled cell images. The app is built with Flask and uses OpenCV for image handling. The goal is to create a simple, effective diagnostic support tool for educational and research purposes.

Features

• Upload and classify blood cell images in real-time.
• Predicts one of the four classes: eosinophil, lymphocyte, monocyte, neutrophil.
• Displays predictions with confidence score.
• Clean, responsive UI with custom HTML and CSS.
• Works locally on any machine with Python installed.

Setup Instructions

1. Prerequisites

• Python 3.12 or above recommended
• Git (to clone the repo)

2. Clone the Repository

git clone https://github.com/raghavaguptha/Hemato-Vision1.git
cd Hemato-Vision1

3. Create and Activate Virtual Environment

python -m venv venv
# On Windows:
venv\Scripts\activate
# On Mac/Linux:
source venv/bin/activate

4. Install Dependencies

pip install -r requirements.txt

5. Run the Flask App

python app.py

Visit http://127.0.0.1:5000/ in your browser to use the app.

How to Use

1. Launch the app and go to the prediction page.
2. Upload a blood cell image (JPEG/PNG).
3. Click "Predict" and the system will classify the image.
4. The result will display the cell type and prediction confidence.

Project Structure

Hemato-Vision1/
│
├── static/               # CSS, images
│   └── assets/           # Backgrounds, icons, etc.
│
├── templates/            # HTML templates (home.html, result.html)
│
├── Blood_Cell.h5         # Trained TensorFlow model (not included in repo due to size)
├── app.py                # Main Flask application
├── requirements.txt      # Dependency list
└── README.md

Tech Stack

• Backend: Flask
• Frontend: HTML, CSS (Bootstrap)
• Image Processing: OpenCV (cv2)
• ML Framework: TensorFlow / Keras
• Language: Python 3.12+

Model Details

• Model Name: Blood_Cell.h5
• Accuracy: ~92% on validation data
• Trained on publicly available blood cell datasets

Known Issues

• The Blood_Cell.h5 file is too large for GitHub upload (over 100 MB). You must download it separately or compress if needed.
• App only accepts .jpg, .jpeg, .png formats.

Future Improvements

• Add support for mobile uploads.
• Enable batch image prediction.
• Deploy online via Render or Hugging Face Spaces.
• Add a feedback loop for incorrect predictions.

License

All rights reserved by Allen015. Consider adding an open-source license for broader contribution and use.

Credits

• Built using open-source libraries: Flask, TensorFlow, OpenCV
• Special thanks to mentors and peers who supported development.