Algasat provides you a web application to upload your satellite images, and quickly get seaweed biomass information. The Algasat platform has been designed to be easy to use and understand, while minimizing the friction for our users.
We use machine learning to identify patterns in your image in order to predict the seaweed biomass. The model has been trained and tuned specifically to identify and predict the biomass of Ascophyllum nodosum.
With Algasat, getting biomass information from your image is as easy as creating a new job in your dashboard. When creating a job, you have to provide a small amount of information to get started, including:
After creating a job, the satellite image is uploaded and it is transformed by our ETL (Extract-Transform-Load) pipeline, known as the Prediction Pipeline. The pipeline extracts features from the image which helps in our prediction.
These features are then transformed into a format that can be ingested by our machine learning model. Our model is a Convolutional Neural Network (CNN) which is commonly used to identify patterns in images.
Once the prediction is made, the pipeline produces a GeoTIFF image, which displays a heat map representation of the biomass. Along with this image, some biomass related statistics are calculated.
The heat map generated will allow you to easily see areas of low and high seaweed biomass, so that you can target the areas for harvesting. You have the option of downloading this file as a .png or a GeoTiff file. A GeoTiff file is an image file that includes additional spatial (georeferencing) information. The GeoTiff file can be opened in any popular GIS (Geographic information system) tool and can give you more information about the exact biomass at a specific location.
Algasat allows you and your team to easily create and share jobs between members. With Algasat, you can sign up your team, and any jobs created by team members are automatically available to the rest of the team in the dashboard.
Algasat is also designed to allow you to share reports with people outside of the platform. After a job has completed, you have the option to share the report as a .pdf or in print.