hi,
i’m tryng to obtain a large image about center of Alps, so I have downloaded 4 different image and the I make a mosaic.
In this example, I join only two image yes, the images are join, and it seems correct:
but when I zoom into, the result are this:
Any suggestion? I don’t know If I wrong something…
thanks a lot
my guess is that the many bright (snow) areas distort the histogram and make the less brighter pixels look black.
Try to move the maximum of the color coding towards 0.2 for red, green, and blue.
This does not change the values of the image but how the colors are distributed over the value range. You should have better contrasts then in these areas.
thank for the advice, it works very well for the zones without snow… However the first image that I uploaded it was only an example, from this image I will create a NDSI for detect snow/not snow.
So my problem is about the zoom into mosaic image (second image uploaded), with your advice I didn’t obtain any changed…
sorry, I was not clear about that. I haven’t used the Multi-size mosaic so far, but I would first test and compare these two options
Also, it seems that the topography somehow was imputed in the image. Did you select “Orthorectify input products” and if so, with which elevation model? Sentinel-2 is quite well georeferenced so I would not recommend doing this by standard.
i didn’t select “Orthorectify input products”, my window is this:
I noticed that if I use Projection as (UTM/WGS 84) I have the problem that I indicated above. While, if I use Geographic Lat/Lon (as in image), i can zoom into the image and I haven’t any problems ( in both test : mosaic_blend, mosaic_overlay)
So, I will export in another program and then changes the projection or are there any possibility for use UTM?
you could first use WGS84 in the Multi-Size Mosaic and then re-project the result in SNAP to UTM.
But this involves resampling two times, quite unnecessary and potentially reducing the image quality.
you are right… i will work on wgs
thanks Braun
