During image acquisition and processing of cellular micrographs for high throughput automated analysis, one commonly encountered problem is segmentation of multiple overlapping cells. In addition, to further classify the cells based on shape information, segmentation should preserve the boundary structures of the cells. In this paper, we combine the distance transform with the marker based watershed algorithm to effectively segment Saccharomyces Cerevisiae cells, fluorescently labeled at their outermost yeast boundaries. This algorithm uses the adaptive threshold to magnify and outline the cell boundaries, adopts the distance transform to turn the position information of pixels to grey level information, and then relies on the watershed algorithm to realize the search and segmentation of the adhesion border of the cells. For classification, invariant features, such as Hu moments, are extracted from segmented cell boundaries. The invariant features enable recognition of objects from imagery in a manner that is independent of scale, position, and orientation. The algorithm was tested on a set of 20 images and produced an average of 90% segmentation rate, and Hu moments with the use of a Multi-Layer Perceptron demonstrated their efficiency with an average of 95% classification rate. We conclude from this data that the currently designed algorithm is robust for classifying fluorescent yeast cells, and it may potentially be applied to more diverse biological samples.