XylemGrower

API

This operator grows Xylem regions in microscopic sections of woods.

Prerequisite is HS[X]-image (hue, saturation, intensity/value/brightness) as the input and some thresholds to control the growing process and an initial segmentation of xylem regions supplied as a binary image.

All three channels of the hs[x] image are assume to be a byte image and values in the range 0 - 255.

The initial regions are eroded to get seed regions, where we determine the mean-value of the hue and/or saturation and/or I/V/B-channel.

Subsequently the region are grown via region growing. We compare each pixel of the contour of those seed regions with the pixel of the background beside them, either in a 4-neighbourhood or an 8-neighbourhood.

This is repeated until the list of uninspected pixel is not empty. Afterwards we may do some post processing. E.G. calculate gradient informations on the found regions to split regions who contain more then one xylem.

Parameters

Results

As a result the detected Xylem regions are returned as a binary image and a table of features for each Xylem regions is created.

Optionally intermediate results may be returned.


© 2010 - 2014 , Alida/MiToBo Development Team
Website: http://www.informatik.uni-halle.de/{alida,mitobo}
Contact: mitobo@informatik.uni-halle.de