This project utilizes deep learning techniques to detect blood groups from infrared images of hands. By analyzing spectroscopic features, the model classifies blood types with high accuracy.

• Automated blood group classification
• Uses infrared hand images for detection
• Deep learning-based model trained on a labeled dataset
• Flask web application for easy interaction

The dataset consists of infrared images categorized into different blood groups:

/dataset_folder
    /train
        /A Positive
        /A Negative
        /AB Positive
        /AB Negative
        /B Positive
        /B Negative
        /O Positive
    /test
    /validation

1. Clone the repository:

   git clone https://github.com/RAHULPATEL2002/blood-group-detection.git
   cd blood-group-detection

2. Install dependencies:

   pip install -r requirements.txt

3. Run the Flask application:

   python app.py

To train the model, execute:

python train_model.py

The trained model is saved as blood_group_model_vgg16.h5.

1. Start the Flask app and open the web interface.
2. Upload an infrared image of a hand.
3. The model predicts and displays the blood group.

• Python
• TensorFlow/Keras
• OpenCV
• Flask

The model achieved an accuracy of 93% on the test dataset. Below is the confusion matrix:

• Rahul Patel - Developer

Feel free to contribute to this project by submitting issues or pull requests!