Skip to main content

Overview

Before installing Reciclaje AI, ensure your system meets the following requirements for optimal performance of the waste detection and classification system.

Hardware Requirements

Minimum Requirements

  • CPU: Dual-core processor (2.0 GHz or higher)
  • RAM: 4 GB
  • Storage: 2 GB free disk space
  • Webcam: USB or built-in camera for real-time detection
  • GPU: Not required (CPU inference supported)
  • CPU: Quad-core processor (3.0 GHz or higher)
  • RAM: 8 GB or more
  • Storage: 5 GB free disk space (for model files and datasets)
  • Webcam: HD camera (720p or higher) for better detection accuracy
  • GPU: NVIDIA GPU with CUDA support for faster inference
    • CUDA 11.x or higher
    • cuDNN 8.x or higher
    • 4GB+ VRAM recommended
While a GPU is not required, it significantly improves real-time detection performance, especially for high-resolution video streams.

Software Requirements

Operating System

Reciclaje AI is compatible with the following operating systems:
  • Windows: 10 or 11 (64-bit)
  • macOS: 10.15 (Catalina) or later
  • Linux: Ubuntu 18.04+, Debian 10+, or equivalent distributions

Python Version

  • Python 3.8 - 3.11 (recommended: Python 3.10)
Python 3.12+ may have compatibility issues with some dependencies. Stick to Python 3.8-3.11 for best results.

Core Dependencies

The following Python packages are required:
PackageVersionPurpose
ultralyticsLatestYOLOv8 model framework
opencv-python4.5.0+Computer vision and video processing
numpy1.21.0+Numerical computations
Pillow8.0.0+Image processing for GUI
imutils0.5.4+Image manipulation utilities
tkinterBuilt-inGraphical user interface
All dependencies will be automatically installed during the setup process. See the Setup Guide for installation instructions.

Camera Requirements

Supported Cameras

  • USB webcams (UVC-compatible)
  • Built-in laptop cameras
  • External IP cameras (with appropriate streaming configuration)

Camera Specifications

  • Resolution: 640x480 minimum (1280x720 recommended)
  • Frame Rate: 15 FPS minimum (30 FPS recommended)
  • Connection: USB 2.0 or higher
The application is configured to use camera index 0 by default. If you have multiple cameras, you may need to modify the camera index in the source code.

Network Requirements

For Model Download

  • Active internet connection (required for initial setup)
  • Minimum 10 Mbps download speed recommended
  • Access to HuggingFace (huggingface.co)

For Runtime

  • No internet connection required after installation
  • All processing is performed locally

Display Requirements

  • Resolution: 1280x720 minimum (for GUI application)
  • Color Depth: 24-bit or higher

Storage Space Breakdown

  • Python packages: ~1.5 GB
  • YOLOv8 model file: ~6 MB
  • GUI assets: ~5 MB
  • Additional space: 500 MB (for temporary files and logs)

Platform-Specific Notes

Windows

  • Visual C++ Redistributable may be required for OpenCV
  • Windows Defender may flag the application during first run

macOS

  • Command Line Tools for Xcode required
  • Camera permissions must be granted in System Preferences

Linux

  • GTK+ or Qt libraries required for OpenCV GUI support
  • Camera access permissions may need to be configured
On Linux, ensure your user is in the video group for camera access:
sudo usermod -a -G video $USER

Verification

To check your Python version:
python --version
To verify camera availability:
python -c "import cv2; print('Camera OK' if cv2.VideoCapture(0).isOpened() else 'No camera')"

Next Steps

Once you’ve verified your system meets these requirements, proceed to:

Setup Guide

Install Python dependencies and configure your environment

Model Download

Download the pre-trained YOLOv8 model from HuggingFace

Build docs developers (and LLMs) love