I am not sure what you mean by accumulated cholesterol. Filipin binds to unesterified cholesterol but not to cholesteryl esters that are found in lipid droplets. The problem that many people have with filipin is that it photobleaches rapidly. This can be minimized with a neutral density filter in the excitation light path.
I mean, for example, NPC cells. They accumulate cholesterol and filipin shows a nice strong signal. When I apply the same protocol to my cells, we get a very weak (almost none) signal, not localized to the plasma membrane despite preparation in the dark and imaging on the same day. So I wondered what the reason might be...
The filipin labeling of NPC cells is much brighter than normal cells. You'll either need to use a more sensitive camera, or focus on another probe at a different wavelength and then use higher brightness of illumination quickly as you switch filters to catch the filipin before it photobleaches. You are likely to see the endocytic recycling compartment near the center of the cell as a bright spot. The plasma membrane will just look like hazy fluorescence in flat cells like fibroblasts. You should be able to see a labeling of the cell perimeter where two cells touch, especially if they are becoming taller as they approach confluence.
We used filipin staining in oocytes that accumulate cholesterol vs control WT oocytes. The labelling in the latter was rather poor, yet different to the background (you can check the protocol in Yesilaltay at el PNAS 2014)
Aleksandra there no difference in binding affinity of filipin in accumulated pool versus deficient (low density) cholesterol pool. What cells are you using? WT cells normally doesn't have high cholesterol hence you will see low signal. The difference you may be seeing may be due to the actual difference in cholesterol level. In PM cholesterol is distributed as opposed of concentrated in pool in lysosomes in NPC hence you see hazy signal.