[Vm-dev] My work featured at Nature News!
JuanVuletich at zoho.com
Thu May 25 13:22:07 UTC 2017
On 23/05/2017 01:15 p.m., Ben Coman wrote:
> On Tue, May 23, 2017 at 11:47 PM, Juan Vuletich <JuanVuletich at zoho.com
> <mailto:JuanVuletich at zoho.com>> wrote:
> Hi Folks,
> Satellogic was featured today at Nature News!
> I helped design and build the hyperspectral cameras in our
> satellites Fresco and Batata. And I wrote the geometric and
> spectral processing software for that image. This is not
> completely off topic, though: The geometric software (image
> rectification and correction), the most complex part of the
> processing, was written by me in Cuis Smalltalk, and runs in a
> Cuis Smalltalk + OpenCL application.
> Please share my joy today!
> Juan Vuletich
> www.cuis-smalltalk.org <http://www.cuis-smalltalk.org>
> Its great to hear of joy in work. It was interesting to read "The
> company announced in January that it would give researchers free
> access to its 30-metre-resolution hyperspectral data. These span
> optical and near-infrared wavelengths and can help track water
> pollution and oil spills, and monitor the health of forests and crops."
Yes. We hope to make a difference.
> Is your geometric software useful dataset users? Or is it just for
> preprocessing to prepare saleable data?
> Is its available to end-users? Or is it secret sauce?
> cheers -ben
The image rectification and geolocation is used to produce the images
most people can use. The idea is to go from pictures taken from the
point of view of the satellite at the moment they were taken (each with
a different perspective) and turn them into a map-like (or Google Maps
like) geometry. This is especially important to be able to track the
evolution of stuff at some point on Earth: you need images from
different points in time, and they need to be able to "stack" them. You
also need to know exactly where each pixel is located (i.e. latitude and
longitude). End users don't need this software: the images we provide
are already corrected.
There are various ways to do this in the industry. We are using a rather
unusual approach and I have developed some novel techniques, that lets
us do this with far less computational cost that usual. These is
currently proprietary technology. I intend to be able to publish them as
part of a Ph.D. thesis on C.S. in a not too distant future. I talked
about all this at
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Vm-dev