Detect Blobs

Detect Blobs

Summary

"Blobs" are connected regions that stand out from their surroundings, usually in terms of pixel intensity. Such regions, once found, are marked in the output image with sequential numeric labels. Pixels belonging to the first detected blob have a value of one. Zero means background.

Downloads

Detailed description

The Detection Blobs Tool has two modes of operation. The mode is selected by setting the Threshold Type parameter to either static or dynamic thresholding mode.

In the static mode you can set an absolute threshold which the intensity of each pixel in the object must exceed. If the Invert flag is false, the qualified pixels are brighter than the threshold. Setting the flag to true reverses the logic. In the example below, the threshold is set such that pixels with an intensity of 200 or higher are qualified. Notice that in the labels in the lower image have different colors because the image is displayed with a "light colors" color map.

Light pills have been detected using a static gray-level threshold.

Light pills have been detected using a static gray-level threshold.

In the dynamic mode the tool decides the threshold automatically and separately for each blob. You can set the sensitivity with which the tool tries to identify prominent objects. The smaller the sensitivity, the sharper the edge of a region must be in order to qualify as a detected blob. in the example below, the sensitivity is set to a value which qualifies both high and low contrast pills.

Dynamic thresholding detects regions that have prominent edges.

Dynamic thresholding detects regions that have prominent edges.

You can also set the minimum and maximum sizes of the regions which will be qualified as blobs. Blob sizes are calculated in world units using an approximate scaling factor for each blob. If the image has bad perspective distortion, the limits may not be accurate.

Typically the labeled image is connected to a Analyze Blob Geometry Tool.