Cassie

About the project

CASSIE is an open source web tool for automatic shoreline mapping and analysis using satellite imagery (www.cassiengine.com).

It was built in JavaScript, using the Google Earth Engine (GEE) API, and can be applied to any coastal region on Earth where there is a boundary between land and water.

CASSIE uses Landsat and Sentinel-2 satellite imagery, freely available from GEE, and implements automatic shoreline detection using normalized differential water index (NDWI) together with the Otsu image segmentation algorithm.

Satellite-derived shorelines are analyzed for a set of user-defined transects along which various statistical analyzes are performed. Comparisons of CASSIE-derived shorelines and rate of change with state-of-the-art methods show that the products of this tool are sub-pixel accurate. The new concept of big data storage and cloud computing (GEE) platform, combined with a user-friendly interface and high level of automation make CASSIE a complete tool to support a wide variety of studies and applications where the knowledge of the behavior of the line coast is essential.

My involvement

In this project, I worked as a developer, where I developed the web interface, the integration with the Google Earth Engine API and the implementation of the automatic coastline detection algorithm. In addition, I also worked as a researcher, where I developed the methodology for automatic shoreline detection and the statistical analysis of transects.

Extras

During my time at Cassie, I coordinated the development of a version of the project called C-Squeeze, focused on mangrove squeeze detection. The project was approved and financed by Grupo Fundação Boticário, and is currently being developed by the other members of the project.

Technologies

• react
• Redux
• Google Earth Engine