SF NDVI

Comment

Comment by Benjamin Golder on October 8, 2010 at 5:13pm

Well, the normalized difference index is very simple, and involved only simple addition, subtraction, and division using the red and infrared bands (see the link for the formula). But as you can see on the left side of the key, vegetation only falls within a small range of that index, so I drew a curve manually in Rhino, and used that manual curve to create a text file for the reclassification. So it was manual, and bent with control points, not logarithmic. Here's the details on the ArcMap tool I used: Reclassify by ASCII file

Comment by Daniel Caserta Segraves on October 8, 2010 at 4:58pm

Yes, I see- looks like you normalized it to a logarithmic scale?
Reads very clearly. Seems like a powerful tool.

Comment by Benjamin Golder on October 8, 2010 at 4:51pm

Thanks.
@Daniel: the image processing was done using ArcMap, which is not open source, though I'm sure that several other major open source softwares, gvSIG, qGIS, and GRASS are all capable of doing the same thing. Grasshopper itself was used in a way that allowed me to draw a curve in Rhino, and manipulate its control points to output different reclassification tables as text. Using Grasshopper allowed me to more "intuitively" slice up the image's colors in order to better see the differences in vegetation.

Comment by Daniel Caserta Segraves on October 8, 2010 at 3:29pm

Fantastic.
So is the image processing code widely available? Did you port/translate it from another application?

Comment by Riccardo Bovo on October 6, 2010 at 2:46am

wow!!

Comment by ĐỒNG VIẾT NGỌC BẢO on October 5, 2010 at 7:31am

cool ! man