Morphology transformations enable you to analyze and modify structures and shapes in an image. Morphology transformations are non-linear and thus allow you to modify the image selectively while other portions of the image remain unchanged. Typical fields of application are quality assurance or medical image processing.

Morphology transformations are a type of neighborhood transformation: the image is analyzed pixel by pixel and each pixel is transformed depending on its neighbors. Thus the results of a morphology transformations depend on two factors:

• The type of morphology operation: the mathematical function used to evaluate each pixel and its neighbors
• The most common operations from which most others are derived are erosion (remove pixels from a bright structure) and dilution (add pixels to a bright structure).
• The structure element: the shape and size of the neighborhood taken into account for the analysis of each pixel
• The structure element reflects the structures in the image you wish to analyze. For example, to find vertical structures use the Vertical structure element.

You can imagine the structure element as a stencil which is moved over the image pixel by pixel. Only the pixels within in the stencil area are visible and taken into account for the morphology transformation. The applied operation (e.g. erosion or delusion) then determines if and how the central pixel within the stencil area is changed.

In the software you first the specify the morphology operation by selecting the corresponding tool. From the tool you then select the structure element and its size (defined by the number of repetitions the structure element is applied).