Account for the samples radii when computing the bounding box of the morphology

[marwan-abdellah]

Currently the samples radii are not included in the calculations of the bounding boxes of the samples radii. For flat morphology, the value of the depth is zero. Please fix!