Training
You train the tool by calling its train() method. Two overloads are provided, one that includes a shape mask and one without. See the figure below.
The model can be any shape derived from the ccShape class. A large variety of shapes are provided by CVL or you can derive you own shape class if necessary. A simple example of a training model is shown in the figure below.
The rectangle and both circles all represent boundaries in this shape example.
The clientFromImage transform will be used at run time to transform target images to client coordinate space.