Skip to main content
Fig. 3 | BMC Bioinformatics

Fig. 3

From: Segmental HOG: new descriptor for glomerulus detection in kidney microscopy image

Fig. 3

Segmentations of a candidate glomerulus. The sizes of the green dots in Panel (a) and Panel (b) represent the boundary likeliness for the candidate glomerulus shown in Fig. 2a. The value of p i takes a natural number between 1 and n(=22) to represent the location on the i-th line segment. For example, in Panel (a), the fourth point in the 26-th line segment is selected, which is expressed as p 26=4. A zigzag boundary of the glomerulus would be obtained if the points that have the largest boundary likeliness are connected naïvely without constraints, as illustrated in Panel (a). By considering the constraints that |p i −p i+1|≤ς(i=1,…,n−1) and |p m −p 1|≤ς where ς=1, a smooth boundary can be obtained as demonstrated in Panel (b)

Back to article page