Classifier

Objects

Objects are extracted from a binary image representing thee regions of interest of the image. For each extracted object a class and cluster is assigned and metrics are measured. Objects are stored to the resulting database and can be used in further pipeline steps.

The Classifier command is used to convert regions of interest, formally extracted from the image using a segmentation algorithm such as Threshold, into objects, where each object is assigned to an object cluster and object class, and metrics are calculated for each object. Classified objects are stored to the resulting database and can be used in further object-processing steps.

ImageC’s classifier provides several filter methods which can be used to to filter out artefact. All objects which match the filter criteria are assigned to the match class, objects not matching the filter criteria are assigned to the no match. The filter can be set for each threshold class individually. This allows the classifier to assign each ROI of a threshold class to a specific match and no match object class. Use the button to add as many filters as needed. As a rule, one filter for each threshold class is used.

Hint

See section Objects for more detailed information about object metrics.