Public Avalanche Reporting Tool 🏔️🚨

Overview

Web application designed to improve avalanche reporting accuracy and reliability.

Problem Statement

Current public avalanche reporting tools face challenges:

• Complicated user experience
• Inaccurate avalanche mapping
• Potential spam or irrelevant images
• Inconsistent avalanche reporting

Features 🚀

Image Validation

• Spam Detection: Advanced image classification
• Prevents system misuse

Avalanche Classification

• Machine learning prediction of avalanche type
• Supports classification of:
  • No avalanche
  • Slab avalanche
  • Loose avalanche
  • Glide avalanche

Interactive Segmentation

• Precise avalanche area marking in images
• Uses Segment Anything Model (SAM)
• Point-based interaction for region highlighting

Size and Type Selection

• Dropdown menus for:
  • Avalanche Size
  • Avalanche Type

Technical Architecture 🏗️

Backend 💻

• FastAPI server
• Machine learning models for:
  • Spam detection
  • Avalanche classification
• Segment Anything Model

Frontend 🎨

• Next.js React application
• Interactive user interface
• State management for image processing

Key Technologies 🛠️

• Python
• FastAPI
• PyTorch
• OpenCV
• Next.js
• React
• Tailwind CSS

Installation 🔧

Prerequisites 📋

• Python 3.10+
• Node.js 14+
• uv
• npm/yarn

Models 🤖

mkdir -p checkpoints/
wget -P checkpoints/ https://dl.fbaipublicfiles.com/segment_anything_2/092824/sam2.1_hiera_large.pt

Backend Setup 🖥️

cd backend
uv sync
python app_fastapi.py

Frontend Setup 🌐

cd frontend
npm install
npm run dev

Future Improvements 🔮

• Enhanced machine learning models
• More granular avalanche classification
• Improved user guidance
• Geospatial data integration