Skip to main content


Figure 12 | BMC Bioinformatics

Figure 12

From: SHERPA: an image segmentation and outline feature extraction tool for diatoms and other objects

Figure 12

Successful segmentation in the presence of debris and overlapping objects. a)c) Micrographs of Fragilariopsis valves, d)f) segmented shapes (highlighted red) after contour optimization, using d) adaptive thresholding, e) Otsu’s thresholding, f) RATS (σ = 3.0). By application of multiple segmentation methods and contour optimization even problematic objects could be extracted, since often at least one of the methods succeeded, but partly at the expense of contour accuracy (see the small bulges in the object contours).

Back to article page