June 5, 2015
Vricon offers a new way to map the world
Maps in two dimension are a thing of the past – or so say the engineers behind Vricon, a new 3-D mapping tool.
Developed as a joint venture between Saab Group, a Swedish aerospace and defense company, and DigitalGlobe, a purveyor of high resolution space and satellite imagery, Vricon uses automated algorithm to process DigitalGlobe’s massive archive of maps into highly accurate, data rich 3D models.
“The core of our technology is called stereophotogrammetry. And that’s been around for ages,” Manne Anliot, Vice President of Global Marketing and Sales at Vricon tells TIME. “That just means taking two images and correlating them with each other. What we are doing, which is unprecedented, is to take more than two images. We actually use all the available imagery over any given area.” The algorithm combines multiple images, sometimes in the hundreds, to create a 3D model which, once mapped, can be updated with new satellite imagery as it is collected by DigitalGlobe’s cluster of satellites.
Vricon provides these models to its clients on an interactive visualization platform, similar in look and feel to Google Earth. “The data is streamed from the Vricon cloud or stored locally with small storage requirements. You only need a very lightweight client and it can be a client on even a handheld or a mobile phone,” says Anliot. Vricon’s clients, which currently include the Swedish Armed Forces and the NGA, at that point have in their hands a tool for geospatial analysis.
“One key feature,” says Anliot, “is the line of sight calculation.” By dropping a pin onto a given point on a Vricon model within their visualization platform, a client can determine what the line of sight is from that position. (See the image below where the yellow colorization denotes what line of sight would be from the pin at the center.) This becomes very handy in, for example, combat situations. “You can position snipers in different locations to make sure that you get the proper coverage of the target that you’re looking for,” says Anliot.
“Another is the reverse. If you know enemy firing positions, you can make calculations from those… in order to calculate a safe route for yourself. It can be used to plan your observation posts, how to be in cover, how to have eyes on your target. It can be to plan your entry and egress route.”
Vricon also has the ability to model change over time. By reprocessing data that has already been mapped, they can automatically detect changes between two data sets, making it possible to detect large scale changes such as receding glaciers, or more recently, the change in the depth and volume of a construction site being dug out at the Uranium enrichment plant in Natanz, Iran between 2010 and 2012.
The current delivery time, from the moment Vricon receives a request from a client to the point of delivery, depends largely on how long it takes to collect the satellite imagery. Once the images are in hand, depending on the size of the area to be mapped, processing can take from a few minutes to a few hours.
“We did Ar-Raqqah for example in Syria, that took us four hours of processing time, from the time we had the imagery until we had the processed product, ” says Anliot. “It took us four hours with our current server farm. But this is completely scalable with the available hardware. Part of this joint venture is that we will scale up our server park and personnel to be able to achieve a production capacity of roughly two million square kilometers per month by mid 2016.”
The company aims to have one third of the globe available as an on-the-shelf streaming product within the next three years. The ultimate goal is to map the entire globe.
“Our focus, the coming years, will be to fully realize the potential of the defense and intelligence needs. And after that it will be infrastructure projects. But I think as we produce the globe, actually having it on the shelf available on our visualization platform, it will open up a lot of very interesting applications.”
No comments:
Post a Comment