Abstract: Fuzzy sets theory is usually applied to represent uncertain, vague objects that cannot be treated as random. Nevertheless, some spatial objects with uncertain boundaries can be characterized by their membership functions that represent probability of the membership. Fuzzy sets operations (conjunction, disjunction, complement) can be uniquely defined for such objects and effective fuzzy algorithms can be utilized accordingly. Results of the fuzzy algorithms have probabilistic meaning, so that standard statistical tests can be used to prove them objectively. This approach has several promising applications, namely land cover classification, cadastral mapping, material quality analysis, radar interferometry.

Supervisor: Lubomír Soukup