*More images and details on this later after some sanitization.
TLDR: Find out if a house is within a new RF coverage area with high resolution, for free.
Most open-source tools don’t have clear integration solutions baked in from one software to another. This is especially true in RF research. When trying to explore a new area, a common path is first uploading known tower and antenna data into Radio Mobile [VE2DBE] to obtain propagation data. This gives the first problem, a .png file. Next, this file can be uploaded into Google Earth (more common due to easy jump in use) or QGIS, which is hopefully correctly mapped. Then finally performing intersection analysis can be automated in QGIS, but this is where the problem with the .png comes in.
A .png file in QGIS is not a true raster and therefore spits out a bunch of odd values, so you have to perform some tricks during analysis to get the results you actually need. Below are some rough steps until I can make a more pretty guide on how to go about this.
drag and drop KML (cannot import kmz to QGIS) into QGIS window
Go to Layer > processing toolbox
>> sample raster > single or batch process > input centroids (points) to check > rf layer (your .png raster file) to check against > output column [r0] for finding centroid overlap > run
This is the point where we have to do some weird tricky stuff. Let me know if you have any issues with this.
>> select new point file > open attribute table > enable edit > select features by value tool, use r0=0 > invert selection (this captures nulls) > delete
>> copy and paste remaining points to the preferred file, if doing multiple rasters
If running separate layers, you will need to cut and paste points from each section, then eliminate duplicate lat long, unless you added another identity column beforehand.
Algorithm fodder. png to raster in qgis, find points in a png image in qgis, find intersects with a image in qgis
boydo.tech 