object classification: vague description of neighborhood features

[thorbenk]

The following paragraph needs to be made clearer. In particular, how do the three neighborhood spin boxes relate to the distance transform? And how many features are added for a single ticked checkbox?

You will also notice features, which can be computed “in the neighborhood”. In that case, the neighborhood of the object ( specified by the user at the bottom of the dialog) is found by distance transform and the feature is computed for the object itself and for the neighborhood including and excluding the object.