How to perform viewshed analysis for each pixel in raster?

Question

I am having a problem making a viewshed analysis for every pixel on a raster. To perform the viewshed analysis from every single pixel in a raster I will need:

• A viewshed plugin (which I am willing to use either the one from GRASS or QGIS plugin for visibility analysis

This is what've done - STEP 1: Get the points layer from the raster

This is my raster: What I've done was used Raster Pixels to Points tool: and then I got this layer: which I think it's was I need for step 2.

• STEP 2: Perform Viewshed Analysis for each Point

As I said I would use one of those plugins/tools, but since the one use in GRASS requires to select one pixel at a time, I was trying to used the other one (QGIS plugin for visibility analysis). So here is the configuration of the tool that I used:

--> From this point on I have problem achieving what I wanted <--

After I run it I have an error associated with the fact that I am not passing some fields ( which I think are also fundamental, observ_hgt and radius) and I don't know either where to put them...

I even tried to add the fields observ_hgt with the value 1.5 and radius 2000 to each pixel (by clicking Toggle Editing > Open Field Calculator) but afterwards I got another error (as can be seen below):

Answer 1

When you click on the Help button in the lower right, you get some, but not really clear informations.

I think you need to add those fields observ_hgt and radius from your error message to your point dataset. The already present id column holds the pixel value, which is your hight value of the DEM at this point. This one has no relevance for the observer and can be deleted.

The help page also states

Only projected raster data can be used. Latitude/longitude “projections”, such as WGS84 are not allowed.

From your screenshot it is clear that your DEM is in EPSG 4236, so indeed unprojected. So you need to reproject your DEM to a projected CRS in a first step, then create your points from it and then add the two columns with meaningful values.

Maybe there are also memory constraints due to the size of your DEM, so better make first tests with a limited selection of points to perform the analysis.