WOLFRAM

Wolfram Language & System Documentation Center

Analyze Segmented Cells in an Image

Use blob analysis techniques on a segmented image to mark the location and size of complete and nonoverlapping cells in a microscopic image.

    
Wolfram Language code: i = [image];
Separate the foreground from the background using binary thresholding and filling the holes.
Wolfram Language code: cells = FillingTransform@Binarize[i, {0, .7}]
Create a marker image by computing the local maxima of the distance transform map.
Wolfram Language code: markers = MaxDetect[DistanceTransform[cells], .25]
Compute the watershed segmentation of the cells using the estimated set of markers and select cells based on their area.
Wolfram Language code: segments = SelectComponents[WatershedComponents[GradientFilter[cells, 2], markers], "Area", 200 < # < 2000&]; Colorize[segments]
Highlight segmented cells using circles of equivalent disk radius.
Wolfram Language code: circles = ComponentMeasurements[segments, {"Centroid", "EquivalentDiskRadius"}]; Show[i, Graphics[{Red, Thick, Circle@@#& /@ circles[[All, 2]]}]]