“…Applications include motion planning for polygonal objects in the presence of polygonal obstacles. Later, these concepts have been generalized to arbitrary shapes in the plane and in space, see (Lee et al, 1998a(Lee et al, , 1998bKaul and Farouki, 1995;Mühlthaler and Pottmann, 2003;Peternell and Manhart, 2003), where the notion of the convolution of two (not necessarily convex) objects has been introduced. 1 Now we consider two regular surfaces A and B in three-dimensional space, which are given by parametric representations a(u, v) and b(s, t) with parameter domains (u, v) ∈ Ω A ⊆ R 2 and (s, t) ∈ Ω B ⊆ R 2 , respectively.…”