Image interpretations of slums contain uncertainties which may negatively impact the accuracy of classification.
We study deviations in slum identification and their delineations as observed from images of three cities.
Existential and extensional uncertainties of slums are modelled in terms of random sets.
Factors of the built environment that experts use to distinguish slums from non-slum areas are identified.
Factors leading to deviations in slum identification and hence the reasons for uncertainties are explored.