Stereo Scape

🌟 StereoScape - 3D Model Generator

Stereo-Scape is a 3D scene reconstruction project that uses stereo vision techniques to generate depth maps and point clouds from pairs of stereo images. It aims to provide a simple, modular, and extensible framework for experimenting with stereo image processing, disparity estimation, and 3D model generation.

🎯 Overview StereoScape leverages cutting-edge Neural Radiance Fields (NeRF) technology to:

Convert 2D images/videos into detailed 3D models Generate high-resolution outputs (up to 4k) Provide realistic rendering with accurate lighting and shadows Enable novel view synthesis for immersive exploration

🚀 Key Features Intuitive Upload Interface: Simple drag-and-drop functionality for images/videos Real-time Processing: Watch your 3D model come to life with progress tracking High-Quality Output: Generate detailed 3D models with realistic textures Interactive Viewing: Explore your 3D models from any angle Fluid Animations: Smooth user experience with beautiful visual effects PSNR Tracking: Real-time Peak Signal-to-Noise Ratio monitoring Multi-view Processing: Support for multiple image angles

💻 Technology Stack Frontend SvelteKit TypeScript Tailwind CSS WebGL animations Backend Python Flask COLMAP TensorFlow MongoDB

🛠️ Getting Started Clone the repository:

git clone https://github.com/nilay2004/steroscape.git

Backend Setup:

cd stereo-scape/backend python -m venv venv source venv/bin/activate # On Windows, use venv\Scripts\activate pip install -r requirements.txt python app.py

Frontend Setup:

cd ../web npm install npm run dev