B E E O D A

Boston Education in Earth Observation Data Analysis



About

BEEODA is a suite of open-source software and educational materials for processing and analyzing earth observation data. It is based on a Linux Ubuntu Virtual Machine that contains a range of open-source image processing tools such as QGIS, Orfeo, GDAL, Python, R, etc. BEEODA is developed and maintained by Pontus Olofsson, Chris Holden and Eric Bullock who are all affiliated with the Department of Earth & Environment, Boston University. To explore BEEODA, please visit our GitHub repository.


Services

We offer to teach workshops using the BEEODA software and materials for analyzing earth observation data anywhere in the world. Workshop topics may include satellite image classification of land-cover, object-based classification of high-resolution imagery, detection and mapping of land-cover change in satellite imagery, time series analysis of satellite data, accuracy assessment of maps and area estimation of mapped land cover and change.

Open Source Modules

Image Pre-Processing

Downloading, atmospheric correction, etc

Classification

Random forest, etc

Object-Based Analysis

Segmentation and Classification

Change Detection

Direct Classification, PCA, MAD, etc

Time Series Analysis

CCDC, YATSM, etc

Area and Accuracy Estimation

Statistical inference


Near-Real Time Alerts

Our research group is in the process of a rigorous comparison of existing and operational near-real time alert systems with our own newly developed ones. Our two algorithms use daily MODIS data to alert of forest disturbances quickly and accurately.


Contact

If you are interested in having us teach a workshop at your institution using the BEEODA material, please contact us at beeodaRS@gmail.com.


Pontus Olofsson

Pontus got his PhD from Lund University, Sweden, in 2007 after studies in physical geography and mathematics and is currently a Research Assistant Professor and Lecturer in the Department of Earth & Environment at Boston University. His research interests include time series analysis of satellite data for land change mapping, terrestrial carbon dynamics and statistical inference of map information. For more information, view Pontus's CV and Google Scholar Profile. Email: olofsson@bu.edu.


Eric Bullock

Eric received his Masters in GIS and Remote Sensing from Boston University in 2014. He is currently a PhD student at BU studying time series analysis of land cover using satellite data. Email: bullocke@bu.edu.


Chris Holden

Chris received his Masters in GIS and Remote Sensing from Boston University in 2011. As a PhD student at BU, his research interests include improving methods for monitoring land cover change and land cover dynamics, the role of land cover change in the carbon cycle, and time series analysis and visualization. Email: ceholden@bu.edu