4-D Visualization of Minkowski Quaternionic Point Set Operations
Jakub Řada, Daniela Velichová, Michal Zamboj · 2026
Read 4-D Visualization of Minkowski Quaternionic Point Set Operations by Jakub Řada, Daniela Velichová, Michal Zamboj (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-25759.txt.
Introduction
Řada, Velichová, and Zamboj visualize Minkowski point set operations in four dimensions. Here the Minkowski product is taken as the quaternionic product. They draw selected point sets with two projection tools. One is double orthogonal projection, and the other is perspective projection, both from 4D to 3D. The sets they generate include circles, which form a Clifford torus and a 3-sphere. Other sets give lines that form a quadratic cone. Some sets contain both circles and lines.
Catalog record
| CREATOR | Jakub Řada, Daniela Velichová, Michal Zamboj |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19354 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.25759
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
4-D Visualization of Minkowski Quaternionic Point Set Operations Jakub Řada, Daniela Velichová, Michal Zamboj Abstract The contribution emphasizes the geometric modeling point of view on Minkowski point set operations. In this paper, the Minkowski product is specified as the quaternionic product. Selected point sets are visualized using double orthogonal projection and perspective projection from four-dimensional to three-dimensional space. In particular, we demonstrate the generation of sets containing circles (Clifford torus, 3-sphere), lines (quadratic cone), or both. MSC2020: Primary 51M15, Secondary 51M04, 00A05. Keywords: Minkowski operations, quaternionic product, four-dimensional visualization, Clifford torus, Hopf sphere 1 Introduction Minkowski point set operations provide a natural way to make geometric objects obey algebraic rules. Two fundamental operations are the Minkowski sum \oplus and the Minkowski product \otimes , defined for points represented by their position vectors. Let A and B be two point sets in \mathbb{R}^{n} . Then A\oplus B:=\{\,a+b\mid a\in A,\;b\in B\,\}, (1) A\otimes B:=\{\,ab\mid a\in A,\;b\in B\,\}. (2) The Minkowski sum was introduced in Minkowski’s papers at the turn of the twentieth century, see, e.g., [ 10 , 11 ] . The Minkowski product was studied more recently and requires a precise specification of the underlying multiplication. In this paper, we restrict ourselves to four-dimensional space \mathbb{R}^{4} , where the Minkowski product is naturally interpreted as the quaternionic product. Our aim is to apply Minkowski set operations to selected parametrically defined subsets of \mathbb{R}^{4} and to visualize the results interactively using 4-D perspective and double orthogonal projections. A modern definition of the Minkowski sum can be found, for example, in [ 17 , p. 126] . The Minkowski sum has applications in various fields, e.g., motion planning [ 9 ] and geometric design [ 12 ] . Quaternionic Minkowski point set operations for unit quaternion sets are described and visualized using stereographic projection in [ 4 ] . Related approaches based on the complex number product in \mathbb{R}^{2} can be found, for example, in [ 2 , 3 , 8 ] . Three-dimensional sets involving the wedge product are discussed in [ 20 , 21 ] , and more generally in [ 22 , 23 ] . Quaternions have been closely connected with their geometric interpretation since their introduction by Hamilton [ 5 , 6 ] . A comprehensive textbook on quaternions with an emphasis on computer visualization is [ 7 ] . An overview of the use of the quaternionic product in \mathbb{R}^{4} in kinematics is given in [ 13 ] . For recent applications to the construction of minimal surfaces from a polynomial perspective, see [ 1 ] . Since the action of unit quaternions induces rotations and provides a parametrization of the 3-sphere, they are often visualized using conformal stereographic projection [ 16 ] . Minkowski quaternionic operations on circles on a unit 3-sphere in stereographic projection also proved to be essential for finding surfaces in \mathbb{R}^{3} that contain two circles through each point [ 18 ] . In our approach, quaternions represent parametrically defined sets (curves, 2-surfaces, and 3-surfaces) in \mathbb{R}^{4} . For the demonstration of various properties, we therefore employ more intuitive methods for visualizing 4-D objects in 3-D: double orthogonal projection, a generalization of Monge’s projection [ 24 , 25 ] , and 4-D perspective [ 14 , 27 ] . Our visualizations, created in Wolfram Mathematica , are interactive 3-D models that often include dynamic manipulators for selected parameters. 1.1 Minkowski quaternionic point set operations in \mathbb{R}^{4} We briefly summarize the necessary preliminaries on quaternions from the four-dimensional point of view adopted in this paper. Let A(a_{0},a_{1},a_{2},a_{3}) and B(b_{0},b_{1},b_{2},b_{3}) be two quaternions considered as points in \mathbb{R}^{4} . The sum A+B of the quaternions A and B is defined componentwise by A+B=\bigl(a_{0}+b_{0},\;a_{1}+b_{1},\;a_{2}+b_{2},\;a_{3}+b_{3}\bigr). (3) The product AB of the quaternions A and B is given by \begin{split}AB=\bigl(a_{0}b_{0}-a_{1}b_{1}-a_{2}b_{2}-a_{3}b_{3},\;a_{0}b_{1}+a_{1}b_{0}+a_{2}b_{3}-a_{3}b_{2},\;\\ a_{0}b_{2}-a_{1}b_{3}+a_{2}b_{0}+a_{3}b_{1},\;a_{0}b_{3}+a_{1}b_{2}-a_{2}b_{1}+a_{3}b_{0}\bigr),\end{split} (4) or, equivalently, in scalar–vector form, \begin{split}AB&=(a_{0},\mathbf{a})(b_{0},\mathbf{b})\\ &=\left(a_{0}b_{0}-\mathbf{a}\cdot\mathbf{b},\;a_{0}\mathbf{b}+b_{0}\mathbf{a}+\mathbf{a}\times\mathbf{b}\right),\end{split} (5) where \mathbf{a}=(a_{1},a_{2},a_{3}) and \mathbf{b}=(b_{1},b_{2},b_{3}) denote the vector parts of A and B , respectively. The neutral element of multiplication is the quaternion 1=(1,0,0,0) .Since quaternion multiplication is non-commutative, we distinguish between left and right products. For quaternions A and B , the left product of B by A is AB and the right product of B by A is BA . Each quaternion A has a conjugate A^{\ast}=(a_{0},-a_{1},-a_{2},-a_{3}). The norm of a quaternion A is defined by |A|=\sqrt{AA^{\ast}}=\sqrt{a_{0}^{2}+a_{1}^{2}+a_{2}^{2}+a_{3}^{2}}. A quaternion U satisfying |U|=1 is called a unit quaternion. For each quaternion A with |A|\neq 0 , the multiplicative inverse is given by A^{-1}=\frac{A^{\ast}}{|A|^{2}}, so that AA^{-1}=1 . Because multiplication is non-commutative, for quaternions A with |A|\neq 0 and B we distinguish between the left division A^{-1}B and the right division BA^{-1} of B by A . Similarly to complex numbers, a quaternion A admits a trigonometric form A=|A|\left(\cos\frac{\varphi}{2}+\frac{\mathbf{a}}{|\mathbf{a}|}\,\sin\frac{\varphi}{2}\right),\qquad\varphi\in[-\pi,\pi], provided that \mathbf{a}\neq\mathbf{0} . From now on, we interpret a quaternion A as the position vector of a point A (keeping the same notation) in \mathbb{R}^{4} . The sum A+B of two points coincides with the classical definition from analytic geometry in \mathbb{R}^{2} or \mathbb{R}^{3} . The quaternionic product AB can be interpreted as a rotation composed with a scaling of B by A , or vice-versa, in \mathbb{R}^{4} , analogously to the complex product in the Gaussian plane \mathbb{C}\cong\mathbb{R}^{2} . Note that points A and B may also be regarded as 0-dimensional point sets in ( 1 ) and ( 2 ). Hence, A+B=A\oplus B and AB=A\otimes B . By introducing parameters, we extend our point sets from individual points to curves, surfaces, and higher-dimensional sets. In general, let A(u_{1},\dots,u_{s}) and B(v_{1},\dots,v_{r}) be parametrically defined point sets of dimensions s and r , respectively, including the special case of 0 -dimensional point sets A and B (without parameters). Their Minkowski sum A\oplus B and product A\otimes B then form point sets of dimension r+s , provided that the embedding space has sufficiently high dimension. In this paper, we restrict ourselves to point sets of dimension at most three in \mathbb{R}^{4} . For example, consider two one-dimensional subsets a(u) and b(v) , represented as curve segments with distinct parameters u and v , given by their normalized parametric representations \begin{split}a(u)&=\bigl(a_{0}(u),a_{1}(u),a_{2}(u),a_{3}(u)\bigr),\qquad u\in[0,1],\\ b(v)&=\bigl(b_{0}(v),b_{1}(v),b_{2}(v),b_{3}(v)\bigr),\qquad v\in[0,1].\end{split} (6) The Minkowski sum a(u)\oplus b(v) is a two-dimensional set — a surface patch in \mathbb{R}^{4} — defined parametrically by q(u,v) over the unit square [0,1]\times[0,1] : q(u,v)=\bigl(a_{0}(u)+b_{0}(v),\;a_{1}(u)+b_{1}(v),\;a_{2}(u)+b_{2}(v),\;a_{3}(u)+b_{3}(v)\bigr). (7) The Minkowski difference a(u)\ominus b(v) is defined analogously. The Minkowski product a(u)\otimes b(v) is again a two-dimensional set p(u,v) defined over [0,1]\times[0,1] : \begin{split}p(u,v)=\bigl(&a_{0}(u)b_{0}(v)-a_{1}(u)b_{1}(v)-a_{2}(u)b_{2}(v)-a_{3}(u)b_{3}(v),\\ &a_{0}(u)b_{1}(v)+a_{1}(u)b_{0}(v)+a_{2}(u)b_{3}(v)-a_{3}(u)b_{2}(v),\\ &a_{0}(u)b_{2}(v)-a_{1}(u)b_{3}(v)+a_{2}(u)b_{0}(v)+a_{3}(u)b_{1}(v),\\ &a_{0}(u)b_{3}(v)+a_{1}(u)b_{2}(v)-a_{2}(u)b_{1}(v)+a_{3}(u)b_{0}(v)\bigr).\end{split} (8) Similarly, the Minkowski left and right divisions are given by a(u)^{-1}\otimes b(v) and b(v)\otimes a(u)^{-1} , respectively. If the point sets A and B share one or more parameters, these may be used to eliminate certain variables. In the case of multiplication (and so division), such a situation leads to a problem of factorization, which lies beyond the scope of this paper. 1.2 4-D to 3-D methods of projection The visualizations presented throughout the paper are constructed using two projection methods from \mathbb{R}^{4} to \mathbb{R}^{3} : double orthogonal projection of the 4-space (x,y,z,w) onto two mutually perpendicular 3-spaces (x,y,z) and (x,y,w) (DOP), and central projection from a projection center in \mathbb{R}^{4} onto a modeling 3-space (x,y,w) (4-D perspective) (Figure 1 ).