[GRASS-dev] Implementation of the High Pass Filter Additive Fusion technique (i.fusion.hpf)

[Moritz Lennert]

On 15/11/13 10:50, Nikos Alexandris wrote:
> Nikos Alexandris wrote:
>
>  > > together with Nikos Ves, we share the "i.fusion.hpf" idea/proof of
>
>  > > concept. At the moment, we have a custom shell script named
>
>  > > `i.fusion.hpf` (an attempt for a proper GRASS add-on), which implements
>
>  > > the High Pass Filter Additive (HPFA) Fusion Technique for
> Pan-Sharpening
>
>  > > [*]. Nikos V started already porting to Python. How can we proceed in
>
>  > > sharing it?
>
>  >
>
>  > [...]
>
>  >
>
>  > > Two questions
>
>  > >
>
>  > > ? Can someone confirm that the part of the existing "i.pansharpen" code
>
>  > > that performs histogram matching (code lines 348 - 431), do so as
>
>  > > "linearly stretching an image to match another image's Mean and
> StdDev"?
>
> Moritz Lennert:
>
>  > AFAICT, it applies the method described in [1].
>
> Is that a reference also indicated in i.pansharpen's manual?

It's not in the manual, but there's a long list of other references. But 
Michael is the one who knows where the inspiration came from. AFAIK, 
this is the classical, generic method of histogram matching.

>
>  > I don't know (and don't have the time to think about) what this
> method does
>
>  > in terms of mean and stddev.
>
> My guess was/is that it is not the same, i.e. it does not match Mean and
> StdDev. As a quick test, I tried the identical (me thinks) tool in
> WhiteboxGIS "Histogram Matching (Two Images)", does not give identical
> Means and StdDevs after the operation -- which is the case with
> i.pansharpen too if I am not wrong.
>

I just did a quick test:

pan in:

mean: 31.813
standard deviation: 3.75447

ms in:

mean: 15.2307
standard deviation: 3.55858

pan out:

mean: 15.6117
standard deviation: 3.23408

So for this example, mean seems to have been adjusted, but stddev not.

>  > > ? Would it be desired to get the HPFA algorithm integrated in
>
>  > > i.pansharpen?
>
>  >
>
>  > Yes. I think that if we have a generic module such as i.pansharpen, it
>
>  > would be preferable to have all pansharpening methods in that one module.
>
> There is one "difference" in that HPFA treats all bands to be sharpened
> separately. And, in this manner, it can be (mis-)used to sharpen any
> low-res band. For example, WorldView-2 products have 8 multi-spectral
> bands. Hence the "not red= green= blue=" design so far from my side.

i.pansharpen does not imply rgb either (although the description of the 
ms* parameters does suggest that. You can obviously use any ms bands you 
want.

Moritz