Almost. The algorithm is actually processing these images and doing feature extraction. You start with completely random arrays, then mess around with them, and you get a brand new image in which the components are separated (ie. forest shows up on one band, river on the other and etc..)

So I’m distributing the processing algorithm. Reading these images is basically just the initial step.

Well, I know this comment is a “bit” late but, I liked your thesis. It’s quite interesting. So, (if I understood ok) your thesis is about how to do distribute the process of reading and “mounting” of ultrahigh-def hyperspectral images, by finding a way to split the algorithm you mention into different processes or threads. Did I caught your thesis? or did I understand potatoes where you said apples?.

[quote comment=”4241″]Congrats. It must feel good; one more year for me..[/quote]

It still haven’t sinked in. I want to make sure I can graduate in May first – then I can feel relieved. :)

Congrats. It must feel good; one more year for me..

Congratulations!