Projects > Media Annotation > Satellite Imagery
Satellite Imagery
Satellite Imagery Technology
The Satellite Imagery project aims at distinguishing certain characteristics of a satellite image. After clicking on one image, the system sorts through the database for similar images. These images are segmented into color coded blocks identifying the type of land that is in that area such as cropland, urban, residential, mountainous areas, or bodies of water.
The military currently has thousands of satellite images from around the world but what are they supposed to do with them all? People hired to filter these images end up quitting because it is such tedious and time consuming. Since there is such a large quantity of satellite images, the manipulation of this data is better suited for computers. The military and other security agencies may be able to use this system to flag potentially hostile territories. For example, new buildings that may be warehouses or nuclear plants could be flagged for further investigation. After the system is able to narrow the location, people may be able to look at more detailed images of this area to base further actions.
Environmentalists may also find this technology to be useful in their line of work. From the categorized land, environmental changes over spans of time may be deduced. In the case of depleting rainforests, this data may help in the fight to restore these lands. Other surveys may be done on land to recognize patterns in occupation or for future planning.
Project Heads
This project is being conducted by James Wang and Jia Li of The Pennsylvania State University.
Demonstration
This is a link to the demonstration page.