Forestnet

We are Providing solution to stop forest fires.

Comment 1

Problem

Wildfires are uncontrolled blazes fueled by weather, wind, and dry underbrush; they can burn acres of land and consume everything in their paths—in mere minutes. Forest fires also contribute to Global warming since many poisonous gases and carbon dioxide are evolved out, which further leads to health issues.

Wildfires and volcanic activities affected 6.2 million people between 1998-2017 with 2400 attributable deaths worldwide from suffocation, injuries, and burns, but the size and frequency of wildfires are growing due to climate change. Hotter and drier conditions are drying out ecosystems and increasing the risk of wildfires.

What it does

A user enters key information about their area, which is then used to predict the probability of a forest fire. This allows them to prepare based on the magnitude and put out time estimations, reducing the impact these natural disasters will have on the environment. Users can also navigate through the index page to find more statistics about wildfires and our mission.

  • LOCALIZING FIRES 1. Clearly delineates the boundaries of the fire. 2. Can be integrated into the video to show progression. 3. Provides a confidence estimator of the localization.

  • PREDICTING MAGNITUDE OF FIRES

    1. Averages several models' output for a final prediction.
    2. Uses GridSearch to tune the best hyperparameters.
    3. Surpasses previous basic regression methods for wildfires.

  • Visualizing Predicted Fires

    1. Interactive map that various features can easily sort
    2. Markers differ in size and color based on fire size
    3. Allows for accurate resource allocation

  • PREDICTING FIRE PUT-OUT TIME

    1. Uses a 3 layered densely connected neural network for predictions.
    2. Model is compiled with Adam optimizer and mean squared error loss.
    3. It Makes it more convenient for a firefighter to allocate resources.

Challenges we ran into

I'm working alone and I don't have much experience in machine learning

  • major challenge is connecting the machine learning model with web
  • I don't have much experience in web development that's why I'm struggling but bootstrap helps me a lot to build websites quickly
Share this project:

Updates