The BioMassters Data Science Competition

The goal of this DrivenData challenge, sponsored by MathWorks, is to predict the quantity of above-ground biomass (AGBM) in a specific region when provided satellite imagery of that region. AGBM is used by farmers, scientists, policymakers, and corporations to understand a forest's carbon stock and to make land use decisions, but it can be time- or resource-intensive to calculate. If this process were streamlined, it would allow these stakeholders to identify patterns and continuously make more informed decisions.

Towards this goal, participants will be expected to develop and implement an algorithm that, based only on the provided imagery and biomass measurements, can predict the peak quantity of biomass for any area in a given year. The contributor who ranks the highest with a MATLAB submission will be awarded a $2000 prize.

Complimentary Software

MathWorks provides complimentary software for this competition. If your team is participating in this competition and needs software, fill out the software request form.

The period to request software has closed.

Student Tutorials and Videos

Learn how to use MATLAB and Simulink to design algorithms, create simulations, deploy code, and speed up software development for your projects.

Watch videos

Competition Resources

MATLAB example code: Estimating Biomass with Deep Learning

General Resources

Preprocess Images for Deep Learning

Prepare Datastore for Image-to-Image Regression

Convert Classification Network into Regression Network

Image-to-Image Regression in Deep Network Designer

Long Short-Term Memory Networks

Image Aggregation of the Sentinel 1A images in MATLAB

Videos

Image Processing, Computer Vision and Deep Learning for Aerial and Satellite Images (41:09)

Image Processing and Computer Vision with MATLAB (19:01)

How to Call Python from MATLAB (3:04)

Using MATLAB with Python (30:49)

Using MATLAB with Tensorflow and PyTorch for Deep Learning

Video Tutorials for Your Team

Interactive Online Tutorials

Learn by doing! Our two-hour online courses provide a hands-on learning environment where you will interact with a web-based version of MATLAB and receive instant and contextual feedback.

View all tutorials