The Performance of Fuzzy Operators on Fuzzy Classification of Urban Land Covers

Abstract

The research discussed in this paper evaluates the performance of selected fuzzy operators (e.g., maximum, minimum, algebraic sum, algebraic product, and gamma operators) for integrating fuzzy membership values associated with multiple spectral bands for mapping the complex spatial mixture that characterises urban land covers. Accordingly, a supervised classification approach based on the fuzzy c-means algorithm was implemented to generate fuzzy memberships of selected bands (1, 3, 4 and 7) of a Landsat-7 ETM image that provided the highest spectral separability among different urban land covers (e.g., forest, grassland, urban, and dense urban) as determined by a transformed divergence analysis. Maps resulting from the application of each fuzzy operator were evaluated against field data. The results show that the fuzzy algebraic product and the fuzzy gamma operators (0.1 to 0.8) are optimal for integrating the fuzzy memberships of selected urban land covers on multi-band data sets, as they exhibited a Khat statistic of 75 percent as compared to a Khat statistic of 59 percent, 64 percent and 71 percent for maximum, minimum and fuzzy algebraic sum, respectively.