Today it is difficult to discuss geospatial image processing without weighing the benefits of cloud computing (read: ENVI in the Cloud). I suspect the same is true in discussing all types of big data analytics, especially in the federal defense and intelligence space. How should decision makers prioritize the analytics they move to the cloud? I believe imagery applications are good candidates to go first. Here’s why…
Geospatial imagery can be cumbersome from a data volume standpoint. It is more efficient to move the image processing capability closer to the data than to move the data closer to each individual end user. As the demand for near real-time imagery derived GEOINT increases from field deployed tactical consumers, access to more efficient cloud-based image processing and analysis becomes essential. After all, it is typically faster to download analysis results that were processed on the cloud than it is to bring the imagery down and conduct the analysis locally.
I acknowledge the concern about moving classified and ITAR controlled data into cloud architectures for obvious national and industrial security reasons. However, there are many analysis tasks that can be performed using unclassified data. Commercial data providers are producing a vast volume of unclassified imagery and providing it in the cloud. New sensors from these data providers will continue to drive data volumes higher making storage a larger and larger concern. Did you hear that Worldview-3 will collect 16 band data for areas as large as 680,000 sq. km per day?!!
In the same way that commercial data providers are providing cloud accessible data, commercial software companies are making geospatial processing capabilities available on the cloud. For example, the ENVI image analysis software contains hundreds of image processing routines that have long been proven on the desktop and recently have been made available in virtual environments. Through the ENVI Services Engine (ESE) these routines can quickly be exposed as applications in the cloud.
With offerings like ENVI Services Engine, Esri ArcGIS Online, and DigitalGlobe cloud access becoming more common place, how fast will you promote imagery applications to the cloud?