Skip to content

chandlerbing65nm/Onboard-Satellite-Image-Classification

Repository files navigation

Image classification of satellite images onboard a nanosatellite.

Contents

personal

Introduction

I designed the software used to classify Earth, Space and Sun-Flare images onboard a nanosatellite. Its purpose is to optimize the bandwidth used in sending the data to the ground station during downlink.

Installation

Requirements

  • Linux/Windows/Mac
  • Python 3
  • Google Colab

Dependencies

pip install -r requirements.txt

Getting-Started

Training and Inference

Run the model.py or model.ipynb to start the training and testing process. The result will give output files of ICU_model.h5 and a lite version model_tflite.

python model.py

Demo

Earth

alt text

Sun Flare

alt text

Space

alt text

License

This repo is licensed under MIT License

About

Image classification of satellite images onboard a nanosatellite.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published