Help Classify Images of Hurricane Sandy Damage

Caitlin Dempsey

FIrst published:

Updated on:

Hurricane Sandy crowdsourced image classification effort.

Volunteers are needed to classify images of Hurricane Sandy to help FEMA.  The site is using open source image sorting tool MapMill, created from code developed  Jeffrey Warren of the Public Laboratory for Open Technology (PLOT) and Science and modified by Kate Chapman (the Acting Executive Director for Humanitarian OpenStreetMap) and Schuyler Erle (one of the authors of O’Reilly’s Mapping Hacks and also of Humanitarian OpenStreetMap).

Upon viewing each image, volunteers are asked to make a determination if the image shows light, moderate, or heavy damage to buildings and infrastructure.  The descriptions for each category:

  • Green: Light or No Damage: shingles or doors missing.
  • Yellow: Moderate Damage: moderate to heavy damage to one or more structures, roads, and infrastructure. Some structures may be only lightly damaged.
  • Red: Severe Damage: heavy damage to multiple structures, or if the area is still flooded.

The site is set up for rapid image evaluation with a basic three category determination.  The site is set up for keyboard shortcuts so hitting ‘1’ will select light, ‘2’ for moderate, ‘3’ for severe, and ‘0’ for a new window.

All of the images from Hurricane Sandy can be browsed here.  Over 300 volunteers have helped categorize over 560 images to date.  The collective results from the crowdsourcing effort will be used to help the Civil Air Patrol (CAP) sort the images.  The images were captured of New York, New Jersey, Rhode Island, and Massachusetts by CAP on October 31, 2012.  Public input on sorting images will help FEMA allocate its resources to the worst damaged areas affected by Hurricane Sandy.

Visit: http://sandy.hotosm.org to start sorting images.  Even helping to categorize just one or two images will help in the collective effort.  Also, be sure to pass along information about this effort to help spread the word.  If you want to select images for a specific area, lick on the grid link to access a gridded map.  Click on the grid for the area you are interested in viewing images for.





Hurricane Sandy crowdsourced image classification effort.
Hurricane Sandy crowdsourced image classification effort.
Photo of author
Caitlin Dempsey
Caitlin Dempsey is a geographer, writer, and the founder and editor of Geography Realm. She holds bachelor's and master's degrees in Geography from the University of California, Los Angeles (UCLA), as well as a Master of Library and Information Science (MLIS) from San José State University. She has been writing about geography, maps, geographic information systems (GIS), and environmental topics for more than two decades through Geography Realm and its predecessor site, GIS Lounge. Her interests include cartography, remote sensing, environmental geography, and the relationship between people and place.