🧿 Iris Image Segmentation Using FCN

This project demonstrates Semantic segmentation of the iris present in the eye images using the Fully Convolutional Network (FCN). The U-Net Architecture is implemented that originates from FCN. The U-Net plays a crucial role in biomedical image segmentation. It is widely used in medical imaging for detection, segmentation and identification. When labelled data is limited, U-Net can learn from fewer training examples as compared with other CNN architectures. The design of the project can be utilised to build applications such as biometrics, iris recognition systems, eye disease detection.

Important Points

• Used OpenCV and TensorFlow/Keras for preprocessing and model building
• Trained and evaluated the model on real eye datasets
• Implemented FCN-based U-Net architecture for iris segmentation
• Visualised the predicted and ground truth images

🛠️ Tech Stack

• Python
• TensorFlow / Keras
• NumPy, Matplotlib, OpenCV
• Jupyter Notebook

🧪 Results

• Achieved 99.85% accuracy on validation dataset with U-Net
• Output Image shows effective separation of the iris region

📂 Repository Structure

🚀 How to Run

1. Clone the repo:

   git clone https://github.com/sfansaria/Iris-Image-Segmentation-Using-FCN.git cd Iris-Image-Segmentation-Using-FCN

2. pip install -r requirements.txt