Mapping Different Types of Shorelines from Coarse-Resolution Imagery: Fuzzy Classification Method Can Deliver Greater Accuracy

Abstract

Hossain, M.S.; Muslim, A.M.; Pour, A.B.; Mohamad, M.N.; Alam, S.M.R.; Nadzri, M.I., and Khalil, I., 2021. Mapping different types of shorelines from coarse-resolution imagery: Fuzzy classification method can deliver greater accuracy. Journal of Coastal Research, 37(2), 433–441. Coconut Creek (Florida), ISSN 0749-0208.Coastal zones are among the most structurally complex ecosystems, though their complex shorelines are threatened because of both natural and anthropocentric influences. There are a smaller number of studies that dealt with developing remote-sensing techniques for detecting and mapping different shoreline types (ST) using coarse-resolution imagery. This study examined fuzzy c-means (FCM), wavelet interpolation, and fuzzy maximum likelihood to map shorelines over Seberang Takir (Malaysia) for different STs. These three fuzzy classification methods were applied to simulated IKONOS (with 16- and 32-m spatial resolutions) image covering the four STs. The positional accuracy of shorelines was assessed in terms of root mean square error (RMSE). The visual inspection and RMSE values showed that variations in accuracies were evident, predominantly due to differences in STs; fuzzy algorithm improved the accuracy. FCM can predict shoreline positions with greater accuracy than the other two methods.