Multi-temporal image analysis for monitoring the changes in fuzzy shorelines

The mapping of shorelines and monitoring of their changes is challenging due to the large variation in shoreline position related to seasonal and tidal patterns. This study focused on a flood-prone area in the north of Java. Shoreline mapping and monitoring are needed to support efforts to recover the eroded land and to reduce the rapid shoreline degradation. We show the possibility of using fuzzy-crisp objects to derive shoreline positions as the transition zone between the classes water and non-water. Fuzzy c-means classification (FCM) was used to estimate the membership of pixels to these classes. A shoreline is represented by a transition zone between the classes, and its spatial extent was estimated by using fuzzy-crisp objects. We analyse the differences between successive shorelines from multi-temporal images. A change category was defined if the membership difference between two years T₁ and T₂, represented by change magnitude values, differs from zero, whereas a no-change category corresponded to a magnitude equal to zero. This resulted into an overall change magnitude and the change directions of the shoreline. It allowed us to jointly identify the trend of the fluctuating shoreline and the uncertainty distribution. Fuzziness of the positions of the shoreline was assessed by using change confusion values. The overall accuracies of FCM resulted into a fuzzy error matrix (FERM) with values ranging from 0.84 to 0.90. Change areas were identified as well as the magnitude, direction and confusion values of the changes for the observed objects at the pixel level. The proposed method provided a way to analyse changes of shorelines as fuzzy objects and is well-suited to apply to coastal areas around the globe.