This project presents a non-invasive approach to blood group prediction using fingerprint image processing and machine learning. By leveraging Convolutional Neural Networks (CNNs), the system classifies fingerprint patterns into eight common blood groups: A+, A-, B+, B-, AB+, AB-, O+, and O-. It aims to provide a quick, accessible, and cost-effective alternative to traditional blood testing methods.
Video Demo: https://youtu.be/BCwa5xclfk0?si=_o926diqvEMfQuql
- Rapid blood group identification with minimal processing time.
- Improved accessibility in remote or resource-limited areas.
- Compatibility with portable and point-of-care diagnostic devices.
- Reduced contamination risk through non-invasive methods.
- Integration of biometric analysis with medical diagnostics using machine learning.
- HTML
- CSS
- JavaScript
- Flask
- SQLAlchemy
- SQLite
- TensorFlow / Keras
- Google Colab
| Model | Testing Accuracy | Validation Accuracy |
|---|---|---|
| VGG16 | 88.72% | 89.50% |
| AlexNet | 12.47% | 12.49% |
| ResNet50 | 61.19% | 62.70% |
| Hybrid (EfficientNetB0 + SVM) | 22.29% | 22.81% |
Source: https://www.kaggle.com/datasets/rajumavinmar/finger-print-based-blood-group-dataset
The dataset consists of fingerprint images labeled with corresponding blood groups. It is used for training, validation, and testing of the deep learning models.
- Authentication and login interface
- Fingerprint upload module
- Blood group prediction result page
- Expansion of the dataset to improve generalization.
- Evaluation of advanced deep learning architectures for higher accuracy.
- Deployment of the model in a production or cloud environment.
Tushar Shinde Email: tusharshinde2250@gmail.com LinkedIn: https://www.linkedin.com/in/tushar-shinde-262335257/
Anjali Maske Email: aamaske50@gmail.com LinkedIn: https://www.linkedin.com/in/anjali-maske/
Contributions are welcome. Feel free to fork the repository, open issues, or submit pull requests. If you find this project useful, consider starring the repository.