I have a raster map consisting of 22 land cover classes (LCC). Each pixel is 300 x 300 meters in dimension and contain one of the 22 LCC. I need to make 1 km x 1 km cells in the parent grid and count the fraction of each land cover class present in that grid.
So I am thinking, in whichever 1x1 km grid the center of each 300x300 m pixel lies, we assign that pixel to the corresponding grid. So, as shown in the figure, if I want to count the fraction of LCC 1 present in the 1x1 km grid, I simply add the number of pixels present in the grid and divide it by the total number of pixels present in the grid. So for LCC 1, we have 3/9 as the fractional count in the drawn grid. The fraction of LCC 16 is 1/9, and so on.
I want to perform this task in QGIS to which I am very new.
I would probably also need to worry about the fact that for 3000m, 6000m, and similar distances, the boundary of large pixel will pass from the middle of the small pixels. In that case how would we decide the location of small pixels?