Phenomaniacs| Automated Detection of Hazards

Automated Detection of Hazards

Countless phenomena such as floods, fires, and algae blooms routinely impact ecosystems, economies, and human safety. Your challenge is to use satellite data to create a machine learning model that detects a specific phenomenon and build an interface that not only displays the detected phenomenon, but also layers it alongside ancillary data to help researchers and decision-makers better understand its impacts and scope.

Where there's smoke...

Summary

Our project has two parts an API which uses machine learning to scan satellite images, and look for smoke as a potential sign of a wildfire, and a front end web page, which allows the user to select an are of the world map (from the a Google Maps view) to be scanned by the API. The results of the machine learning query are then returned to the user along with the satellite image that was scanned.

How We Addressed This Challenge
  • We built a machine learning model which is trained to detect smoke in a satellite image, and an API through which to access this model. We also built a front end web page using HTML, CSS, and Vanilla Javascript which as a user interface to consume the API. The two together will check for signs of a wildfire in that part of the earth.
  • This system could provide the beginnings of and eventual AI model that could continually scan satellite imagery giving early warnings of wildfires.
  • This system allows a user to select an area of the world map using a Google Maps views. This is then sent to our API which makes use of the NASA World View API and machine learning to detect whether there is smoke found in a satellite image of that part of the globe. The conclusion of this as well as the satellite image which was looked at is then returned to the user's view.
  • Our hope would be that this solution could be further developed into a system that automatically detects and gives early warning of wildfires.
How We Developed This Project

We chose this challenge because it was just that: a challenge. None of us knew machine learning prior to this project. We began by building the machine learning model and then moved on to the front end that would consume it. We built the machine learning model using ML.Net, and then built a C#.NET API through which it can be accessed. Our front end is built using HTML, CSS, and Vanilla Javascript. It also makes use of the Google Maps API as well as the NASA World View API. Our biggest achievement in this project was the successful building and training of a machine learning model to recognize smoke when given a satellite image, and even distinguish between smoke and clouds.

How We Used Space Agency Data in This Project

We made extensive use of the NASA World View API for satellite imagery to train the model with, as well as to feed to the model for scanning.

Project Demo
Project Code
Data & Resources
  • NASA World View API
  • Google Maps API
  • ML.NET
Tags
#MachineLearning, #AutomaticDetectionOfHazards, #Fire
Judging
This project was submitted for consideration during the Space Apps Judging process.