Module: ClassifyPixels

ClassifyPixels classify image pixels as belonging to different classes using the machine-learning tool, ilastik.

ClassifyPixels performs per-pixel classification using the ilastik application. Ilastik is now bundled with the CellProfiler distribution; it applies supervised machine learning techniques to images to learn their features. A user trains a classifier with Ilastik and then saves the classifier. The user then uses the ClassifyPixels module to classify the pixels in an image.

ClassifyPixels produces an "image" consisting of probabilities that the pixel belongs to the chosen class; this image is similar to an intensity image that would be produced by fluorescence imaging. Provided that the classifier is sufficiently accurate, the image is well-suited for input into one of the Identify modules for object detection. More instructions on using the interface may be found here. Please note that you must use the same image format for classification as for the initial learning phase.

Currently, ilastik is only available for Windows, and is accessible from in the CellProfiler folder under the Start Menu. A 64-bit system is recommended for running ilastik.

Settings:

Classifier file location

Select the folder containing the classifier file to be loaded. You can choose among the following options which are common to all file input/output modules:

Elsewhere and the two sub-folder options all require you to enter an additional path name. You can use an absolute path (such as "C:\imagedir\image.tif" on a PC) or a relative path to specify the file location relative to a directory):

Classfier file name

This is the name of the Classfier file.

Select the class

Select the class you want to use. The class number corresponds to the label-class in ilastik