Semi-supervised fuzzy C-means clustering for change detection from multispectral satellite image

Data clustering has been applied in almost areas such as health, natural resource management, urban planning... especially, fuzzy clustering which the advantage with handling better for ambiguous data. This paper proposes a method of improving fuzzy c-means clustering algorithm by using the criteria to move the prototype of clusters to the expected centroids which are pre-determined on the basis of samples. The proposed algorithm is used for a model of change detection on multispectral satellite imagery at multiple temporals. The experiments are implemented on various data sets in comparison with other approaches.