Existence of Anisotropic Minkowski Content
Filip Fryš · 2025
Read Existence of Anisotropic Minkowski Content by Filip Fryš (2025) 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-2508-08156.txt.
Introduction
Filip Fryš presents a 2025 arXiv paper on anisotropic Minkowski content, a boundary measure that depends on direction. For a set of finite perimeter, he asks when the tilted versions of this measure exist and what they equal. The result is a tidy equivalence: the ordinary boundary content matches the perimeter exactly when the anisotropic content matches the average of the set's and its complement's anisotropic perimeters. A consequence covers outer contents as well.
Catalog record
| CREATOR | Filip Fryš |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 63 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-08156
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Existence of Anisotropic Minkowski Content Filip Fryš Abstract . This paper is devoted to the existence of anisotropic Minkowski content and anisotropic outer Minkowski content. Our result is that the Minkowski content of the topological boundary of a given set of finite perimeter E E coincides with the perimeter of E E if and only if the anisotropic Minkowski content of the topological boundary of E E coincides with half of the sum of the anisotropic perimeter of E E and the anisotropic perimeter of the complement of E E . As a consequence, we find that the existence of anisotropic outer Minkowski content of a given set of finite perimeter and its complement ensures the existence of outer Minkowski content of the set and its complement. Introduction Let \emptyset\neq S\subseteq\mathbb{R}^{n} be measurable. We define the Minkowski content of S as \mathcal{M}(S;\mathbb{R}^{n})\coloneq\lim_{\varepsilon\to 0_{+}}\frac{\lambda^{n}\big(\{x\in\mathbb{R}^{n};\;\dist(x,S)<\varepsilon\}\big)}{2\varepsilon}, if the limit exists, where \lambda^{n} stands for the n -dimensional Lebesgue measure and \dist(\cdot,S) for the distance function to the set S . It turns out that the numerator of the fraction in the definition of \mathcal{M}(S;\mathbb{R}^{n}) can be replaced by the measure of the Minkowski sum S\oplus B(0,\varepsilon)\coloneq\bigcup_{s\in S}B(s,\varepsilon), and so \mathcal{M}(S;\mathbb{R}^{n})=\lim_{\varepsilon\to 0_{+}}\frac{\lambda^{n}\big(S\oplus B(0,\varepsilon)\big)}{2\varepsilon}, if the limit on the right hand side exists. If the set S is the boundary of a compact, convex and nonempty set, i.e. the boundary of a convex body , then the limit exists and coincides with the (n-1) -dimensional Hausdorff measure (see [ Sch , Chapter 5] ). A natural question that arises is whether the limit, if it exits, coincides with (n-1) -dimensional Hausdorff measure for more general sets. Indeed, if the set S is compact and (n-1) -rectifiable (i.e. the image of a compact subset of \mathbb{R}^{n-1} under a Lipschitz map), we find that the limit \mathcal{M}(S;\mathbb{R}^{n}) exists and coincides with the (n-1) -dimensional Hausdorff measure [ Fe , p. 275] , or, more generally, if the set S is compact, countably \mathcal{H}^{n-1} -rectifiable (see Definition 1.1 ) and there exists a probability Radon measure \eta absolutely continuous with respect to \mathcal{H}^{n-1} and a positive constant \gamma such that for any x\in S and any r\in(0,1) we have \eta\big(B(x,r)\big)\geq\gamma r^{n-1}. Its proof can be found in [ AFP , Theorem 2.104] . It is also possible to restrict ourselves to an open subset \Omega\subseteq\mathbb{R}^{n} and define the Minkowski content of S in \Omega by (1) \mathcal{M}(S;\Omega)\coloneq\lim_{\varepsilon\to 0_{+}}\frac{\lambda^{n}\Big(\big((S\cap\Omega)\oplus B(0,\varepsilon)\big)\cap\Omega\Big)}{2\varepsilon}, if the limit exists. One could ask whether we can reasonably replace the set B(0,1) in ( 1 ). This idea leads to the notion of anisotropic Minkowski content: Let C\subseteq\mathbb{R}^{n} be a convex body the interior of which contains the origin. We define the C -anisotropic Minkowski content of S by \mathcal{M}_{C}(S;\Omega)\coloneq\lim_{\varepsilon\to 0_{+}}\frac{\lambda^{n}\Big(\big((S\cap\Omega)\oplus B(0,\varepsilon)\big)\cap\Omega\Big)}{2\varepsilon} if the limit exists, where (S\cap\Omega)\oplus\varepsilon C\coloneq\{s+\varepsilon c;\;(s,c)\in(S\cap\Omega)\times C\}. In the C -anisotropic case, we will compare the C -anisotropic Minkowski content of the topological boundary of a set of finite perimeter E in an open set \Omega with the value \frac{1}{2}\int\limits_{\Omega\cap\partial^{*}E}\big(h_{C}(\nu_{E})+h_{C}(-\nu_{E})\big)\difm^{n-1}, where \nu_{E}(x) is the so-called reduced boundary of E and \nu_{E}(x) is the outer unit normal to E at x\in\partial^{*}E (see Definition 1.4 in this text) and h_{C} is the support function of C given by h_{C}(x)\coloneq\sup_{y\in C}x\cdot y\quad\text{for $x\in\mathbb{R}^{n}.$} See [ LV , Theorem 3.1] for an analogue of [ AFP , Theorem 2.104] for C -anisotropic Minkowski content. In this work, however, we are mainly interested in the role of C : We will show in Theorem 3.1 that the Minkowski content of the topological boundary of a given set of finite perimeter exists and coincides with the perimeter of E , i.e. \mathcal{M}(\partial E;\Omega)=\mathcal{H}^{n-1}(\Omega\cap\partial^{*}E) if and only if the C -anisotropic Minkowski content of the topological boundary of a given set of finite perimeter exists and coincides with the arithmetic mean of the C -anisotropic perimeter of E and the C -anisotropic perimeter of the complement of E , i.e. \mathcal{M}_{C}(\partial E;\Omega)=\frac{1}{2}\int\limits_{\Omega\cap\partial^{*}E}\big(h_{C}(\nu_{E})+h_{C}(-\nu_{E})\big)\difm^{n-1}. Its proof of this statement is based on the Besicovitch covering theorem and [ LV , Lemma 3.3] . The authors of [ CLL ] introduced the C -anisotropic outer Minkowski content of a given set E of finite perimeter in an open set \Omega as \mathcal{SM}_{C}(E;\Omega)\coloneq\lim_{\varepsilon\to 0_{+}}\frac{1}{\varepsilon}\lambda^{n}\Big(\big((E\cap\Omega)\oplus\varepsilon C)\big)\cap(\Omega\setminus E)\Big), if the limit exists. Its isotropic version, i.e. for C being the unit Euclidean ball, was introduced in [ ACV ] . For a convex set K , the value of \mathcal{SM}_{B(0,1)}(K;\mathbb{R}^{n}) coincides with the (n-1) -dimensional Hausdorff measure of the boundary of K and \mathcal{SM}_{C}(K;\mathbb{R}^{n}) coincides with \int\limits_{\partial K}h_{C}(\nu_{K})\difm^{n-1}, see again [ Sch , Chapter 5] . In the case of anisotropic outer Minkowski content of E in \Omega , we will compare its value with \int\limits_{\Omega\cap\partial^{*}E}h_{C}(\nu_{E})\difm^{n-1}. Our result—Theorem 3.1 —is an analogue of the following theorem [ CLL , Theorem 3.4] : Theorem ( [ CLL ] ) . Let \Omega\subseteq\mathbb{R}^{n} be open, E\subseteq\Omega be a set of finite perimeter in \Omega and C be a convex body the interior of which contains the origin. If \mathcal{SM}_{B(0,1)}(E;\Omega)=\mathcal{H}^{n-1}(\Omega\cap\partial^{*}E), then \mathcal{SM}_{C}(E;\Omega)=\int\limits_{\Omega\cap\partial^{*}E}h_{C}(\nu_{E})\difm^{n-1}. The authors of [ CLL ] assert without proof that the converse should also be true: If \mathcal{SM}_{C}(E;\Omega)=\int\limits_{\Omega\cap\partial^{*}E}h_{C}(\nu_{E})\difm^{n-1} for a sufficiently nice C , then \mathcal{SM}_{B(0,1)}(E;\Omega)=\mathcal{H}^{n-1}(\Omega\cap\partial^{*}E). In [ CLL ] , there is a short comment on how to prove this assertion. Unfortunately, however, it seems that it is not as straightforward as the authors write in [ CLL , Remark 3.5] . We will use our Theorem 3.1 to show a weaker result, namely that if the closure of E is contained in \Omega and \displaystyle\mathcal{SM}_{C}(E;\Omega)=\int\limits_{\Omega\cap\partial^{*}E}h_{C}(\nu_{E})\difm^{n-1} and \displaystyle\mathcal{SM}_{C}(\Omega\setminus E;\Omega)=\int\limits_{\Omega\cap\partial^{*}E}h_{C}(-\nu_{E})\difm^{n-1}, then \displaystyle\mathcal{SM}_{B(0,1)}(E;\Omega)=\mathcal{SM}_{B(0,1)}(\Omega\setminus E;\Omega)=\mathcal{H}^{n-1}(\Omega\cap\partial^{*}E). The rest of this paper is organized as follows. In the first section— 1. Preliminaries —we present the notation that will be used, collect the preliminaries on geometric measure theory and anisotropic (outer) Minkowski content, and lay out all the tools we need in the following sections, particularly sets of finite perimeter , the Besicovitch covering theorem , anisotropic perimeter , anisotropic (outer) Minkowski content and their properties. In the second section— 2. Basic Properties of Anisotropic (Outer) Minkowski Content —we take a closer look at necessary conditions for the existence of anisotropic (outer) Minkowski content and study the relation between anisotropic Minkowski content and anisotropic outer Minkowski content. In the last section— 3. Main Result —we give a proof of Theorem 3.1 which tells us that the anisotropic Minkowski content of the topological boundary of a given set of finite perimeter exists if and only if the isotropic Minkowski content of the topological boundary of that set exists. Furthermore, we study the relation between the anisotropic Minkowski content of the reduced boundary of a given set of finite perimeter and the anisotropic outer Minkowski content of the set of points which have Lebesgue density one with respect to that set (Proposition 3.6 ). 1. Preliminaries 1.1. Notation Let n\in\mathbb{N} (and so n\geq 1 ). For x,y\in\mathbb{R}^{n} , we denote by x\cdot y=\sum_{i=1}^{n}x_{i}y_{i} the Euclidean scalar product of x and y , and the symbol |\cdot| stands for the norm induced by this scalar product. If A,B\subseteq\mathbb{R}^{n} are general and E\subseteq\mathbb{R}^{n} is measurable, then: • The symbol A\oplus B\coloneq\{a+b;\;a\in A,\;b\in B\} stands for the Minkowski sum of A and B . • The