I have a raster with 1,0 values only (pixel 50mx50m). I also have very little experience in working with rasters.....
What I would like to do is possibly smooth the raster to fill in some of the gaps in the clusters (if it is possible).
I would also like to do some type of cluster analysis to pick up the main clusters in the dataset. Ideally I am looking for clusters of 5ha (20 cells). Due to the clusters being fragmented in areas I would like to ideally like to locate buffers of 20 cells within a 2 pixel buffer.
I have added a couple of pictures showing the cluster patterns.
What I have done so far is to convert to polygons and assess the data from there however due to the gaps in the pixels this method isn't picking up the entire cluster.