You might need to smooth even more.
Michelle Umali wrote:
> Dear Bruce & Doug,
> Thanks I'll do the label comparison.
> I am attaching a fieldsign pic to show the irregularity of the
> blue/red bands. Can you suggest a method of determining V1/2/3 based
> on the odd pattern?
>
> Thank you very
Hi Michelle
they have to overlap unless the classification was 100% accurate with no
variability. You can just pick the label with the highest probability at
each vertex
cheers
Bruce
On Tue, 16 Aug 2011, Michelle Umali wrote:
> Hi Doug,
> So I did the additional smoothing and the map looks bet
can you send a pic?
Michelle Umali wrote:
> Hi Doug,
> So I did the additional smoothing and the map looks better.. It's
> still hard to interpret though. I was expecting alternating red and
> blue bands to signify V1/V2/V3 boundaries, but it's very irregular
> with red curves running ever
Hi Doug,
So I did the additional smoothing and the map looks better.. It's
still hard to interpret though. I was expecting alternating red and
blue bands to signify V1/V2/V3 boundaries, but it's very irregular
with red curves running every which way into blue splotches. Do you
have any s
