This project aims to automate the classification and cropping of AADHAR card images using a trained deep learning model. It provides tools to detect, classify, and verify the authenticity of AADHAR cards using PyTorch and Keras.

• 📄 Detect and classify AADHAR card images using a ResNet model.
• ✂️ Crop the relevant sections from the card automatically.
• ✅ Includes average prediction logic for improved accuracy.
• 🧠 Pretrained models included (Card_Detector_ResNet.keras & best.pt).

Screenshot 2025年03月23日 194249

git clone https://github.com/<your-username>/AADHAR-Classification-and-Cropping.git
cd AADHAR-Classification-and-Cropping