mkhontohendry300-stack/Melano-Cancer-Scanner

MelanoScan – AI Skin Cancer Detection Prototype An AI-powered skin cancer screening prototype that detects and classifies skin lesions using deep learning models built in Roboflow. MelanoScan aims to make early cancer screening more accessible, affordable, and efficient—especially in under-resourced communities. The system uses object detection an # Melano-Cancer-Scanner 📌 Overview MelanoScan is an AI-driven prototype designed to detect and classify skin lesions using simple image uploads. The goal is to make early skin cancer screening accessible, affordable, and fast, especially in communities that lack access to dermatology specialists. The system demonstrates how AI + no-code tools (Roboflow) can be used to address global healthcare challenges. Based on the project proposal Week 2 – MelanoScan Cancer Detector Week 2-Proposal for MelanoScan … . 🎯 Project Purpose Skin cancer rates are rising worldwide. Early detection greatly improves treatment success. Many people only seek help when the disease has advanced. Some communities lack dermatologists and diagnostic tools. ➡️ MelanoScan provides a simple AI-powered screening solution using smartphone images. Week 2-Proposal for MelanoScan … ⭐ Key Features 📸 Upload an image of a skin lesion 🔍 Detect lesion region using RF-DETR (Object Detection) 🧬 Classify lesion type using ViT Classification 🏷️ Identifies 7 skin lesion categories (benign & malignant) 📱 Includes a mobile UI prototype (Figma) 🏆 Designed as a low-cost, accessible early-screening tool 🧠 AI Models Used 1️⃣ Object Detection (RF-DETR Small) Locates the lesion in the image. Week 2-Proposal for MelanoScan … 2️⃣ Classification (ViT – Vision Transformer) Classifies the lesion into one of 7 types: Malignant: bcc, akiec, mel Benign: bkl, df, nv, vasc Week 2-Proposal for MelanoScan … 🗂️ Skin Lesion Classes Malignant Lesions BCC – Basal Cell Carcinoma AKIEC – Actinic Keratoses / Intraepithelial Carcinoma MEL – Melanoma Benign Lesions BKL – Benign Keratosis-like Lesions DF – Dermatofibroma NV – Melanocytic Nevi VASC – Vascular Lesions Week 2-Proposal for MelanoScan … 🔄 Workflow User uploads an image of a skin lesion. Model detects the lesion area and crops it. Classification model predicts the lesion type. Output: Image with bounding box Predicted cancer type Confi …