SkinAI — Skin Disease Detection System

Project Overview

SkinAI is a Python-based web application that uses an AI model to help screen common skin lesions from images. It provides a simple and responsive interface where users can upload a skin photo, run inference, and see the most likely skin condition along with risk information and educational guidance.

This project is intended for learning and screening only. It is not a medical diagnosis tool.

What This Project Does

• Loads a trained skin disease classification model from Model/skin_disease_model.keras
• Accepts image uploads through a Flask web interface
• Predicts one of several skin lesion classes using the model
• Displays the predicted condition, confidence, and risk category
• Includes supportive descriptions and next-step guidance for each condition
• Works locally on your machine, with mobile access via QR code if enabled

Key Features

• Flask-powered web dashboard
• Image upload and preview support
• Model inference for skin conditions
• Condition summaries and risk-level display
• Local app deployment with one command
• Clean project structure for training, inference, and web UI

Repository Structure

• app.py — main Flask application and inference endpoint
• train.py — training script for building or re-training the model
• requirements.txt — Python dependencies
• README.md — project description and usage guide
• Model/ — trained model and metadata files
  • skin_disease_model.keras
  • class_names.json
  • training_config.json
• templates/index.html — HTML interface for the web app
• static/css/style.css — frontend styling
• static/js/app.js — client-side behavior and AJAX
• HAM10000_images_part_1/, HAM10000_images_part_2/ — dataset image folders
• HAM10000_metadata.csv — dataset metadata for training

Setup Instructions

1. Open PowerShell in the project root folder.
2. Create and activate a virtual environment:

python -m venv .venv
.\.venv\Scripts\Activate.ps1

3. Install dependencies:

pip install -r requirements.txt

Run the Project Locally

From the project folder, run:

python app.py

Then open the browser at the local address shown in the terminal.

GitHub Push Instructions

If you already have a GitHub remote configured, use these commands from the project root:

git checkout main
git status
git add .
git commit -m "Add SkinAI project files and README"
git push origin main

If you need to set the GitHub remote first, create a repository on GitHub, then run:

git remote add origin https://github.com/<your-username>/<your-repo-name>.git
git branch -M main
git push -u origin main

Before pushing, make sure:

• git status shows the files you want to commit
• the remote URL uses your GitHub account
• large files like model weights are acceptable for your repository size

Important Notes

• Do not treat this app as a medical diagnosis system.
• The model prediction is only a screening suggestion.
• Always verify with a licensed dermatologist for real medical concerns.

Dependencies

This project uses the following Python packages:

• Flask
• streamlit
• qrcode
• tensorflow
• numpy
• pandas
• pillow
• scikit-learn
• matplotlib

License

This project is shared for educational and development use.