/Melanoma-and-Nevus-Segmentation-and-Classification

Primary LanguagePython

Melanoma and Nevus Segmentation and Classification Project

Overview

This project applies advanced machine learning techniques for the segmentation and classification of melanoma and nevus. It's divided into two main parts: CNN classification using ResNet50 and UNet-based segmentation followed by SVM classification.

Goal

The goal is to provide a tool for early detection and diagnosis of skin cancers, aiding healthcare professionals.

Features

  • ResNet50 CNN: Classify skin lesions as melanoma or nevus.
  • UNet Segmentation: Segment skin cancer lesions.
  • SVM Classification: Classify extracted features from segmented images.

Technologies

  • Deep Learning (CNN, UNet)
  • Support Vector Machine (SVM)
  • Image Processing (Color Analysis, Entropy)

License

MIT License

Contact

For more information, please contact [sidhoummohamedcherif@gmail.com].

Replace [sidhoummohamedcherif@gmail.com] with actual contact information.