In this paper, an iterative cell image segmentation algorithm using sh
ort-time Fourier transform magnitude vectors as class features is pres
ented, The cluster centroids of the magnitude vectors are obtained by
the K-means clustering method and used as representative class feature
s. The initial image segmentation classifies only those image pixels w
hose surrounding closely matches a class centroid, The subsequent proc
edure iteratively classifies the remaining image pixels by combining t
heir spatial distance from the regions already segmented and the simil
arities between their corresponding magnitude vectors and the cluster
centroids, Experimental results of the proposed algorithm for segmenti
ng real cell images are provided.