Intersection Properties of Finite Disk Collections

Preprint

Article

Intersection Properties of Finite Disk Collections

Altmetrics

Downloads

Views

Comments

A peer-reviewed article of this preprint also exists.

Jesús F. Espinoza^*,

Cynthia G. Esquer-Pérez

Jesús F. Espinoza^*,

Cynthia G. Esquer-Pérez

This version is not peer-reviewed

Submitted:

11 January 2024

Posted:

15 January 2024

You are already at the latest version

Alerts

Abstract

In this work we study the intersection properties of a finite disk system in the euclidean space. We accomplish this by utilizing subsets of spheres with varying dimensions and analyze specific points within them, referred to as poles. Additionally, we introduce two applications: estimating the common scale factor for the radii that makes the re-scaled disks intersects in a single point, this is the \v{C}ech scale, and constructing the minimal Axis-Aligned Bounding Box (AABB) that encloses the intersection of all disks in the system.

Keywords:

Subject: Computer Science and Mathematics - Computational Mathematics

MSC: 68U05; 65D17; 05E45

1. Introduction

One of the new techniques developed for the analysis of large clusters of information, known as Big Data, is Topological Data Analysis (TDA). In TDA, simplicial complexes associated with the data are constructed. These structures include the Vietoris-Rips complex, the Čech complex, and the piecewise linear lower star complex, among others. Of special interest to us is the generalized Čech complex structure. Although the standard Čech complex is formed by intersecting a collection of disks with a fixed radius, the generalized version allows varying radii. This flexibility enables us to highlight specific data points by assigning or weighting them with larger and/or more rapidly expanding balls to them, while de-emphasizing others by using smaller and/or slower growing balls. This approach proves valuable for handling noisy data sets, offering an alternative to discarding data that may not meet a specific significance threshold [1].

Understanding the patterns of intersections and the timing of intersections among a set of disks in

R^{d}

, each with potentially different radii, is a fundamental problem. This leads to the exploration of the generalized Čech complex structure, which captures the intersection information of these disks, regardless of their radii. Rescaling the radii by the same factor, we obtain a filtered generalized Čech complex, where the associated simplicial complexes evolve as the scale parameter varies. In particular, in [2], algorithms are provided to calculate the generalized Čech complex in

R^{2}

, and [3] presents an algorithm to determine the Čech scale for a collection of disks in the plane.

To establish the necessary foundation for our study, Section 2 introduces crucial concepts and notation that will be used throughout the article and we focus on analyzing the intersection of a disk system in

R^{d}

. We start by investigating the intersection of two disks in Section 2.1 and then expand our analysis to a system of m disks in Section 2.2. By applying Helly’s Theorem, we prove that it is sufficient to examine the intersection of all subsystems consisting of

d + 1

disks in order to determine if the system has a nonempty intersection.

In Section 3, we define Vietoris-Rips systems and Čech systems, together with presenting results regarding the Rips scale and the Čech scale, as well as their connections. In Section 3.1, we present an algorithm that can determine whether the intersection of the system is empty or non-empty. This is achieved by exclusively computing the poles of subsystems of disks (or spheres). Finally, in Section 3.2, we introduce the algorithm that computes an approximation to the Čech scale using the numerical bisection method.

Additionally, in Section 4 we incorporate the concept of an minimal axis-aligned bounding box (AABB) into our methodology. An AABB is a rectangular parallelepiped whose faces are perpendicular to the basis vectors. These bounding boxes frequently arise in spatial subdivision problems, such as ray tracing [4] and collision detection [5]. In this paper, we study AABBs to enclose the intersection of a finite collection of disks. This approach proves valuable for discerning whether the collection intersects at a singular point or not. In this section, we also provide an algorithm for constructing the AABB of a disk system.

2. Intersection properties of sphere systems

Throughout this work, we will refer to a d-disk system M, or simply a disk system, as a finite collection of closed disks in

R^{d}

with positive and not necessarily equal radii, i.e.,

M = {D_{i} (c_{i}; r_{i}) \subset R^{d} ∣ c_{i} \in R^{d}, r_{i} > 0, 1 \leq i \leq m < \infty} .

Moreover, in order to study the intersection properties of a disk system M with the approach addressed in Section 3 and Section 4 of this work, we will conduct a study in this section of the intersection properties of the spheres corresponding to the boundaries of each disk in M, which we call a sphere system and denote by

\partial M

\partial M = {\partial D_{i} \subset R^{d} ∣ D_{i} \in M},

where ∂ denotes the topological boundary operator.

Following the notation in [6], we introduce the following generalization of the sphere.

Definition 1.

An i-sphere in

R^{d}

is the intersection of a sphere with an affine subspace of dimension i.

Of course, the notions of a sphere (as a

(d - 1)

-dimensional surface) and a d-sphere in

R^{d}

agree. However, an i-sphere in

R^{d}

can also be viewed as the intersection of d-spheres. For instance, the intersection of two spheres typically occurs in a hyperplane, forming a

(d - 1)

-sphere in

R^{d}

. When another d-sphere intersects this configuration, the result may be a

(d - 1)

-sphere, a

(d - 2)

-sphere, a 0-sphere (a single point), or it might even be empty, all within the same hyperplane. For a disk system

M = {D_{i} (c_{i}; r_{i})}

composed of m disks, where

{c_{1}, \dots, c_{m}}

is a set in general position in

R^{d}

, the maximum dimension of the affine subspace associated with the i-sphere, obtained from the intersection of all the spheres in

\partial M

, is at most

d - m + 1

, or equivalently,

i = d - m + 1

. This conclusion is drawn from [6, Theorem 2.1] and the fact that the affine hull of

{c_{1}, \dots, c_{m}}

is of dimension

m - 1

. Consequently, the following result is proven.

Lemma 1.

Let

M = {D_{1} (c_{1}; r_{1}), \dots, D_{m} (c_{m}; r_{m})}

be disk system such that

{c_{1}, \dots, c_{m}}

is a set in general position in

R^{d}

. Then, the possibilities for the set

\cap_{D_{i} \in M} \partial D_{i}

are:

the empty set;
a single point;
a $(d - m + 1)$ -sphere.

Remarkable points in i-spheres that will play a key role in the rest of the article are the poles. Let

π_{i} : R^{d} ⟶ R

be the canonical projection on the i-th factor for

i = 1, . . ., d

, and let

{e_{1}, e_{2}, . . ., e_{d}}

be the standard basis of

R^{d}

Definition 2.

Let

e_{q}

be the q-th vector of the canonical base of

R^{d}

. An

e_{q}

-north (south) pole of an i-sphere S in

R^{d}

is a point on S whose projection on the q-th coordinate is maximum (minimum). In other words,

x \in S

is the

e_{q}

-north pole if

π_{q} (y) \leq π_{q} (x)

for all

y \in S - {x}

, where

π_{q}

represents the projection onto the q-th coordinate.

We denote the

e_{q}

-north pole of S by

s_{q}^{+}

and the

e_{q}

-south pole by

s_{q}^{-}

An i-sphere can have a single

e_{q}

-pole (north or south) or an infinite number of them, which occurs when a normal vector to the affine space containing the i-sphere is aligned with the vector

e_{q}

. We are interested in finding the

e_{q}

-poles of

(d - m + 1)

-spheres originating from disk systems

M = {D_{1} (c_{1}; r_{1}), \dots, D_{m} (c_{m}; r_{m})}

, by taking the intersection

\cap_{j = 1}^{m} \partial D_{j}

. Such

(d - m + 1)

-spheres will be denoted by

S_{M} (c; r)

, to emphasize the disk system M, as well as its center and radius.

Lemma 2.

Let

M = {D_{1}, \dots, D_{m}}

be a d-disk system such that

⋂_{j = 1}^{m} D_{j} \neq \emptyset

, and let p be a point in

⋂_{j = 1}^{m} D_{j}

such that

π_{q} (p) \leq π_{q} (x)

(resp.

π_{q} (p) \geq π_{q} (x)

) for every x in

⋂_{j = 1}^{m} D_{j}

. Then, there exists an i-sphere

S = \partial D_{j_{1}} \cap \dots \cap \partial D_{j_{i}}

such that p is in S and p is the

e_{q}

-south pole (resp.

e_{q}

-north pole) of S.

Proof.

Since

\cap_{j = 1}^{m} D_{j} \neq \emptyset

, then

\partial (\cap_{j = 1}^{m} D_{j}) \neq \emptyset

\partial (\cap_{j = 1}^{m} D_{j}) \subset \cap_{j = 1}^{m} D_{j}

due to the closedness of the sets

D_{j}

, for

j = 1, \dots, m

, and

p \in \partial (\cap_{j = 1}^{m} D_{j})

On the other hand, since

\partial (\cap_{j = 1}^{m} D_{j}) \subset \cup_{i = 1}^{m} \partial D_{j}

, there exist indices

j_{1}, \dots, j_{i}

such that

p \in \partial D_{j_{r}}

for any

r = 1, \dots, i

; let

Λ (p) = {j_{1}, \dots, j_{i}} \subseteq {1, \dots, m}

be a maximal subset of indices such that

p \in D_{j}

if and only if

j \in Λ (p)

. We claim that

p \in \cap_{r = 1}^{i} \partial D_{j_{r}}

is the

e_{q}

-south pole of

S : = \cap_{r = 1}^{i} \partial D_{j_{r}}

In effect, let

V_{p} \subset R^{d}

be an open neighborhood of p sufficiently small such that:

Every $x \in V_{p} \cap \partial (\cap_{j = 1}^{m} D_{j})$ has as maximal set of indices a proper subset of $Λ (p)$ ,
$S \cap V_{p} \subset \partial (\cap_{j = 1}^{m} D_{j})$ .

The first condition can be guaranteed by the finiteness of the disk system M, and the second condition is a consequence of the maximality of the set

Λ (p)

. Therefore,

π_{q} (p) \leq π_{q} (x)

for every

x \in S \cap V_{p}

, which is equivalent to the fact that

π_{q} (p) \leq π_{q} (x)

for every

x \in S

, in the case of i-spheres.

□

2.1. Sphere systems with two spheres

In the following two lemmas we provide the computations to determine the center, radius and poles for a

(d - 1)

-sphere given by the intersection of two disks in

R^{d}

Lemma 3.

Let

M = {D_{1} (c_{1}; r_{1}), D_{2} (c_{2}; r_{2})}

be a disk system with two d-disks such that

\partial D_{1} \cap \partial D_{2}

is a

(d - 1)

-sphere

S = S_{M} (c; r)

with center c and radius r. Then,

c = \frac{1}{2} (1 + \frac{r_{2}^{2} - r_{1}^{2}}{∥ c_{2} - c_{1} ∥^{2}}) c_{1} + \frac{1}{2} (1 + \frac{r_{1}^{2} - r_{2}^{2}}{∥ c_{2} - c_{1} ∥^{2}}) c_{2}

and

r = \frac{2 \sqrt{s (s - ∥ c_{2} - c_{1} ∥) (s - r_{1}) (s - r_{2})}}{∥ c_{2} - c_{1} ∥}

where

s = \frac{1}{2} (∥ c_{2} - c_{1} ∥ + r_{1} + r_{2})

Proof.

Let

Π

be the hyperplane containing the

(d - 1)

-sphere S, which is defined by the equation:

\sum_{i = 1}^{d} (k_{i} - h_{i}) x_{i} - \frac{1}{2} \sum_{i = 1}^{d} (k_{i}^{2} - h_{i}^{2}) = \frac{r_{1}^{2} - r_{2}^{2}}{2},

(1)

where

c_{1} = (h_{1}, \dots, h_{d})

and

c_{2} = (k_{1}, \dots, k_{d})

. Then the normal vector of the hyperplane

Π

is given by

N : = c_{2} - c_{1} = (k_{1} - h_{1}, \dots, k_{d} - h_{d})

, and the center c of S is determined by the intersection point of the hyperplane

Π

with the perpendicular line that passes through the center

c_{1}

D_{1}

. This line can be parameterized as

γ : t \mapsto c_{1} + t N = (x_{1} (t), \dots, x_{d} (t))

, such that

γ (0) = c_{1}

and

γ (1) = c_{2}

. We can compute the intersection point

c = γ (t_{*})

Π

and

γ ([0, 1])

, for any

t_{*} \in [0, 1]

, by substituting it in (1),

\sum_{i = 1}^{d} (k_{i} - h_{i}) (h_{i} + t_{*} (k_{i} - h_{i})) = \frac{1}{2} (\sum_{i = 1}^{d} (k_{i}^{2} - h_{i}^{2}) + (r_{1}^{2} - r_{2}^{2})) .

And solving for

t_{*}

, we obtain that

t_{*} = \frac{1}{2} + \frac{r_{1}^{2} - r_{2}^{2}}{2 ∥ c_{2} - c_{1} ∥^{2}}

. Hence, the center of S is given by:

\begin{matrix} c = c_{1} + t_{*} N = \frac{1}{2} (1 + \frac{r_{2}^{2} - r_{1}^{2}}{∥ c_{2} - c_{1} ∥^{2}}) c_{1} + \frac{1}{2} (1 + \frac{r_{1}^{2} - r_{2}^{2}}{∥ c_{2} - c_{1} ∥^{2}}) c_{2} \end{matrix}

Next, we will compute the radius r of S. This radius can be determined as the height r of the triangle with base

∥ c_{2} - c_{1} ∥

formed by the points

c_{1}

c_{2}

, and a point on S. Thus, by the Heron’s formula we have that

r = \frac{2 \sqrt{s (s - ∥ c_{2} - c_{1} ∥) (s - r_{1}) (s - r_{2})}}{∥ c_{2} - c_{1} ∥},

where

s = \frac{1}{2} (∥ c_{2} - c_{1} ∥ + r_{1} + r_{2})

correspond to the semi-perimeter. □

We can proceed now to compute the poles of the

(d - 1)

-sphere

\partial D_{1} \cap \partial D_{2}

Lemma 4.

Let

D_{1} (c_{1}; r_{1})

and

D_{2} (c_{2}; r_{2})

be two d-disks such that

\partial D_{1} \cap \partial D_{2}

is a

(d - 1)

-sphere

S = S (c; r)

with center c and radius r. Then, the

e_{q}

-poles of S are

s_{q}^{\pm} = c \pm \sum_{i}^{d} x_{i} e_{i}

, where

x_{i} = \{\begin{matrix} \frac{r | π_{i} (c_{2} - c_{1}) π_{q} (c_{2} - c_{1}) |}{∥ c_{2} - c_{1} ∥ \sqrt{∥ c_{2} - c_{1} ∥^{2} - π_{q} {(c_{2} - c_{1})}^{2}}}, & i \neq q \\ \frac{r \sqrt{∥ c_{2} - c_{1} ∥^{2} - π_{q} {(c_{2} - c_{1})}^{2}}}{∥ c_{2} - c_{1} ∥}, & i = q . \end{matrix}

Proof.

For simplicity, we translate the hyperplane

Π

, which contains the

(d - 1)

-sphere S, as well as the sphere itself, to the origin; in such case, the corresponding equations are given by,

\begin{matrix} \sum_{i = 1}^{d} (k_{i} - h_{i}) x_{i} & = 0, \\ \sum_{i = 1}^{d} x_{i}^{2} & = r^{2}, \end{matrix}

where

h_{i} : = π_{i} (c_{1})

and

k_{i} : = π_{i} (c_{2})

for

i = 1, \dots, d

. In the case that

k_{q} - h_{q} = π_{q} (c_{2} - c_{1}) = 0

, the normal vector

N = c_{2} - c_{1}

of the hyperplane

Π

is orthogonal to the basis vector

e_{q}

. Therefore, the

e_{q}

-poles of S are

c \pm r e_{q}

, which agree with the formulae of the lemma.

On the other hand, suppose that

k_{q} - h_{q} \neq 0

. To find the

e_{q}

-poles of S, we will use the Lagrange multiplier method. Consider the following function:

x_{q} = f (x_{1}, x_{2}, . . ., \hat{x_{q}}, . . ., x_{d}) = \frac{- \sum_{j \neq q}^{d} (k_{j} - h_{j}) x_{j}}{k_{q} - h_{q}},

(2)

subject to the restriction:

g (x_{1}, x_{2}, . . ., \hat{x_{q}}, . . ., x_{d}) = \sum_{j \neq q}^{d} x_{j}^{2} + {(\frac{- \sum_{j \neq q}^{d} (k_{j} - h_{j}) x_{j}}{k_{q} - h_{q}})}^{2} - r^{2} = 0

Let

λ

be the Lagrange multiplier, we define

h (x_{1}, . . ., \hat{x_{q}}, . . ., x_{d}, λ) = f (x_{1}, . . ., \hat{x_{q}}, . . ., x_{d}) + λ g (x_{1}, . . ., \hat{x_{q}}, . . ., x_{d})

For any

i \neq q

, consider the following system of equations:

\frac{\partial h}{\partial x_{i}} = - \frac{k_{i} - h_{i}}{k_{q} - h_{q}} + 2 λ x_{i} + 2 λ (\frac{- \sum_{j \neq q}^{d} (k_{j} - h_{j}) x_{j}}{k_{q} - h_{q}}) (- \frac{k_{i} - h_{i}}{k_{q} - h_{q}}) = 0 .

Then

\begin{matrix} - \frac{k_{i} - h_{i}}{k_{q} - h_{q}} + 2 λ (x_{i} + \frac{k_{i} - h_{i}}{{(k_{q} - h_{q})}^{2}} \sum_{j \neq q}^{d} (k_{j} - h_{j}) x_{j}) & = 0 \\ - (k_{i} - h_{i}) (k_{q} - h_{q}) + 2 λ ({(k_{q} - h_{q})}^{2} x_{i} + (k_{i} - h_{i}) \sum_{j \neq q}^{d} (k_{j} - h_{j}) x_{j}) & = 0 \end{matrix}

Solving this system of equations for

λ

, we obtain that,

λ = \frac{(k_{i} - h_{i}) (k_{q} - h_{q})}{2 ({(k_{q} - h_{q})}^{2} x_{i} + (k_{i} - h_{i}) \sum_{j \neq q}^{d} (k_{j} - h_{j}) x_{j})}

Comparing the last expression for two indices

i \neq \tilde{i}

, we have that,

\begin{matrix} x_{i}^{2} & = \frac{r^{2} {(k_{i} - h_{i})}^{2}}{{(k_{\tilde{i}} - h_{\tilde{i}})}^{2} + \sum_{j \neq \tilde{i}, q} {(k_{j} - h_{j})}^{2} + \frac{1}{{(k_{q} - h_{q})}^{2}} {(\sum_{j \neq q} {(k_{j} - h_{j})}^{2})}^{2}} \\ = \frac{r^{2} {(k_{i} - h_{i})}^{2}}{\sum_{j \neq q} {(k_{j} - h_{j})}^{2} + \frac{1}{{(k_{q} - h_{q})}^{2}} {(\sum_{j \neq q} {(k_{j} - h_{j})}^{2})}^{2}} \\ = \frac{r^{2} {(k_{i} - h_{i})}^{2} {(k_{q} - h_{q})}^{2}}{({(k_{q} - h_{q})}^{2} + \sum_{j \neq q} {(k_{j} - h_{j})}^{2}) (\sum_{j \neq q} {(k_{j} - h_{j})}^{2})} \\ = \frac{r^{2} π_{i} {(c_{2} - c_{1})}^{2} π_{q} {(c_{2} - c_{1})}^{2}}{∥ c_{2} - c_{1} ∥^{2} (∥ c_{2} - c_{1} ∥^{2} - π_{q} {(c_{2} - c_{1})}^{2})} \end{matrix}

Finally, for

i = q

, we can use the last expression to substitute it in (2) and obtain the desired result. □

2.2. Sphere systems with more than two spheres

Now, let us proceed with the explicit calculation of the coefficients for the center c of the

(d - m + 1)

-sphere

S = \cap_{j = 1}^{m} \partial D_{j}

. We can achieve this by considering the disk system translated to

c_{m}

, denoted as

{D_{j} (c_{j} - c_{m}; r_{j})}_{j = 1}^{m}

, and by defining the

(d - m + 1)

-sphere

S - {c_{m}} = \cap_{j = 1}^{m} \partial D_{j} (c_{j} - c_{m}; r_{j})

. This sphere is positioned at the intersection of hyperplanes (for more details, refer to [6]).

{(c_{k} - c_{m})}^{T} x = \frac{1}{2} (r_{m}^{2} + ∥ c_{k} - c_{m} ∥^{2} - r_{k}^{2})

(3)

for all

k = 1, . . ., m - 1

. Utilizing the information that the center of

S - {c_{m}}

can be expressed as a combination of the centers

c_{k} - c_{m}

and substituting it into (3), we obtain a linear system of equations with dimensions

(m - 1) \times (m - 1)

\sum_{j = 1}^{m - 1} λ_{j} (c_{k} - c_{m}) \cdot (c_{j} - c_{m}) = \frac{1}{2} (r_{m}^{2} + ∥ c_{k} - c_{m} ∥^{2} - r_{k}^{2})

for

k = 1, . . ., m - 1

. Solving the system of equations for

λ = (λ_{1}, . . ., λ_{m - 1})

, we find the center of S as follows:

c = λ_{1} (c_{1} - c_{m}) + . . . + λ_{m - 1} (c_{m - 1} - c_{m}) + c_{m}

The radius of the sphere S can be computed using the equation:

r^{2} = r_{k}^{2} - {∥ c - c_{k} ∥}^{2}

for any

k \in {1, . . ., m - 1}

Now that we have determined the center and radius of S, as well as the affine space that contains it, we can proceed to compute its

e_{q}

-poles for each

q \in {1, 2, . . ., d}

. These poles reside in the affine space that contains S and within a set that we define below.

Let S be an i-sphere in

R^{d}

, and let

n_{1}, . . ., n_{d - i}

be orthogonal vectors to the affine space L that contains S. Consider the space M generated by these vectors together with the vector

e_{q}

from the canonical basis of

R^{d}

. Let us denote

n_{j} = {(n_{l}^{(j)})}_{l = 1}^{d}

for each

j = 1, . . ., d - i

. Then, we can define

L_{0}

, the set L translated to the origin, as follows:

\begin{matrix} L_{0} & = {〈{n_{j}}_{j = 1}^{d - i}〉}^{⊥} \\ = \{x \in R^{d} ∣ x \cdot n_{j} = 0, \forall j = 1, . . ., d - i\} \end{matrix}

The set M is defined as:

\begin{matrix} M & = 〈{n_{j}}_{j = 1}^{d - i} \cup {e_{q}}〉 \\ = \{x = \sum_{j = 1}^{d - i} λ_{j} n_{j} + λ_{d - i + 1} e_{q} \in R^{d} ∣ λ_{j} \in R, \forall j = 1, . . ., d - i + 1\} \\ = \{(\sum_{j = 1}^{d - i} λ_{j} n_{1}^{(j)}, . . ., \sum_{j = 1}^{d - i} λ_{j} n_{q}^{(j)} + λ_{d - i + 1}, . . ., \sum_{j = 1}^{d - i} λ_{j} n_{d}^{(j)}) ∣ λ_{j} \in R\} \end{matrix}

Refer to Figure 1 for a visual representation of the subspaces L and

M + {c}

As mentioned above, the

e_{q}

-poles of S lie at the intersection of L and

M + {c}

, where c is the center of S. To simplify the calculations, we will utilize

L_{0}

and M, and then translate them into c. The intersection of M and

L_{0}

can be expressed as follows:

\begin{matrix} M \cap L_{0} & = \{x \in M ∣ x \cdot n_{k} = 0, \forall k = 1, . . ., d - i\} \\ = \{\sum_{j = 1}^{d - i} λ_{j} n_{j} + λ_{d - i + 1} e_{q} \in R^{d} |(\sum_{j = 1}^{d - i} λ_{j} n_{j} + λ_{d - i + 1} e_{q}) \cdot n_{k} = 0, λ_{j} \in R \forall k = 1, . . ., d - i\} \\ = \{\sum_{j = 1}^{d - i} λ_{j} n_{j} + λ_{d - i + 1} e_{q} \in R^{d} |\sum_{j = 1}^{d - i} λ_{j} n_{j} \cdot n_{k} + λ_{d - i + 1} n_{q}^{k} = 0, λ_{j} \in R \forall k = 1, . . ., d - i\} \end{matrix}

Let us consider a disk system in

R^{d}

, denoted

{D_{j} (c_{j}; r_{j})}_{j = 1}^{m}

, where

m < d

. The intersection of their boundaries forms a

(d - m + 1)

-sphere S. In this case, the subspace M has dimension m, or

d i m (M) = m - 1

e_{q} \in M

. We choose the normal vectors for the affine space containing S as

n_{j} = c_{j} - c_{m}

, where

j = 1, . . ., m - 1

. Then

M = \{(\sum_{j = 1}^{m - 1} λ_{j} (c_{1}^{(j)} - c_{1}^{(m)}), . . ., \sum_{j = 1}^{m - 1} λ_{j} (c_{q}^{(j)} - c_{q}^{(m)}) + λ_{m}, . . ., \sum_{j = 1}^{m - 1} λ_{j} (c_{d}^{(j)} - c_{d}^{(m)})) |λ_{j} \in R\}

By rewriting, we have

\begin{matrix} M \cap L_{0} & = \{\sum_{j = 1}^{m - 1} λ_{j} n_{j} + λ_{m} e_{q} \in R^{d} |\sum_{j = 1}^{m - 1} λ_{j} n_{j} \cdot n_{k} + λ_{m} n_{q}^{k} = 0, λ_{j} \in R, \forall k = 1, . . ., m - 1\} \end{matrix}

S (c; r) = \cap_{j = 1}^{m} \partial D_{j}

is the

(d - m + 1)

-sphere with center in c and radius r, then the

e_{q}

-poles of S are the

e_{q}

-poles of

S - {c_{m}}

but translated by c. The poles of

S - {c_{m}}

are located in

M \cap L_{0}

. If p is an

e_{q}

-pole of

S - {c_{m}}

, then it can be expressed as

p = \sum_{j = 1}^{m - 1} λ_{j} n_{j} + e_{q} λ_{m}

for some

λ_{j} \in R

j = 1, . . ., m

and the following conditions holds:

\sum_{j = 1}^{m - 1} λ_{j} n_{j} \cdot n_{k} + λ_{m} n_{q}^{k} = 0

for each

k = 1, . . ., m - 1

and

{∥ p ∥}^{2} = r^{2}

Thus, if

p = \sum_{j = 1}^{m - 1} λ_{j} n_{j} + e_{q} λ_{m}

is an

e_{q}

-pole of

S - {c_{m}}

, the following equations are satisfied for

λ_{1}, λ_{2}, . . ., λ_{m} \in R

\sum_{j = 1}^{m - 1} λ_{j} n_{j} \cdot n_{k} + λ_{m} n_{q}^{k} = 0

(4)

\sum_{i = 1}^{d} {(\sum_{j = 1}^{m - 1} λ_{j} n_{i}^{j})}^{2} + 2 λ_{m} \sum_{j = 1}^{m - 1} λ_{j} n_{q}^{j} + λ_{m}^{2} = r^{2}

(5)

for all

k = 1, . . ., m - 1

, with r the radius of the

(d - m + 1)

-sphere S. From (4) we have the system

(\begin{matrix} n_{1} \cdot n_{1} & n_{1} \cdot n_{2} & . . . . & n_{1} \cdot n_{m - 1} \\ n_{2} \cdot n_{1} & n_{2} \cdot n_{2} & . . . . & n_{2} \cdot n_{m - 1} \\ . . . \\ n_{m - 1} \cdot n_{1} & n_{m - 1} \cdot n_{2} & . . . . & n_{m - 1} \cdot n_{m - 1} \end{matrix}) (\begin{matrix} λ_{1} \\ λ_{2} \\ . . . \\ λ_{m - 1} \end{matrix}) + λ_{m} (\begin{matrix} n_{q}^{1} \\ n_{q}^{2} \\ . . . \\ n_{q}^{m - 1} \end{matrix}) = 0 .

Let us denote A as the matrix

{(n_{i} \cdot n_{j})}_{i, j}

and B as the vector

{(- n_{q}^{j})}_{j = 1}^{m - 1}

. Then, we have

A λ = λ_{m} B

, where

λ = (λ_{1}, . . ., λ_{m - 1})

. Solving for

λ

, we obtain

λ_{j} = λ_{m} (A^{- 1} B) [j]

for each

j = 1, . . ., m - 1

(where

(A^{- 1} B) [j]

denotes the entry j of the

(m - 1) \times 1

vector

(A^{- 1} B)

). By substituting the value of

λ_{j}

into (5), we obtain the quadratic equation:

λ_{m}^{2} [\sum_{i = 1}^{d} {(\sum_{j = 1}^{m - 1} (A^{- 1} B) [j] n_{i}^{j})}^{2} + 2 \sum_{j = 1}^{m - 1} (A^{- 1} B) [j] n_{q}^{j} + 1] - r^{2} = 0

Let us define

Γ_{i} = \sum_{j = 1}^{m - 1} (A^{- 1} B) [j] n_{i}^{j}

for all

i = 1, . . ., d

. Solving this equation, we find:

λ_{m} = \frac{\pm r}{\sqrt{\sum_{i = 1}^{d} Γ_{i}^{2} + 2 Γ_{q} + 1}}

λ_{j} = \frac{\pm r (A^{- 1} B) [j]}{\sqrt{\sum_{i = 1}^{d} Γ_{i}^{2} + 2 Γ_{q} + 1}}

for

j = 1, . . ., m - 1

. Therefore, the

e_{q}

-poles of S, for

q = 1, . . ., d

, are:

p = \sum_{j = 1}^{m - 1} λ_{j} n_{j} + e_{q} λ_{m} + c = \frac{\pm r}{\sqrt{\sum_{i = 1}^{d} Γ_{i}^{2} + 2 Γ_{q} + 1}} (\sum_{j = 1}^{m - 1} (A^{- 1} B) [j] n_{j} + e_{q}) + c

3. Vietoris-Rips and Čech systems

Our goal in this section is to provide a comprehensive understanding of the disk system, the Vietoris-Rips system, and the Čech system. Additionally, we introduce some results that establish a certain connection between both disk systems. Investigating the features and qualities of data and spaces can provide us with useful knowledge about their geometric and topological characteristics.

Before we look into the definitions of Vietoris-Rips and Čech systems, let us give a brief overview. These systems are essential in the field of topological data analysis for recognizing and comprehending the geometric structure of point cloud data. Vietoris-Rips complex and Čech complex share the goal of capturing the topology of the underlying metric space, both provide different ways of recognizing connections and associations among data points. The Vietoris-Rips complex tends to be more efficient and scalable for large datasets, while the Čech complex can be more accurate but computationally more expensive. The choice between the two depends on the nature of the dataset and the specific goals of the topological analysis. Now, let us move on to defining these fundamental concepts.

Definition 3.

Let

M = {D_{1}, D_{2}, \dots, D_{m}}

be a d-disk system. We say M is a Vietoris-Rips system if

D_{i} \cap D_{j} \neq \emptyset

for each pair

i, j \in {1, 2, \dots, m}

. Furthermore, if the d-disk system M has the nonempty intersection property

⋂_{D_{i} \in M} D_{i} \neq \emptyset

, then M is called a Čech system.

For each

λ \geq 0

, we define a collection of d-disks

M_{λ}

with the same centers as those in the d-disk system M, but with radii rescaled by

λ

. When

λ > 0

M_{λ}

is a d-disk system again.

M_{1}

is equal to M, and

M_{0}

is the set of the centers of the d-disks in M.

In the field of topological data analysis, the Rips scale and the Čech scale are essential parameters for determining the closeness and connectivity between data points. These two scales offer different perspectives on how we measure and comprehend geometric relationships within point-cloud data. To understand their importance in capturing the underlying topological structure, let us look at their definitions. The Vietoris-Rips scale

ν_{M}

, of a d-disk system M is the smallest

λ \in R

such that

M_{λ}

is a Vietoris-Rips system. Similarly, the Čech scale

μ_{M}

, of M is the smallest

λ \in R

such that

M_{λ}

is a Čech system. This is

\begin{matrix} ν_{M} & = inf {λ \in R ∣ M_{λ} i s a V i e t o r i s - R i p s s y s t e m} \\ μ_{M} & = inf {λ \in R ∣ M_{λ} i s a Č e c h s y s t e m} \end{matrix}

Next, we present some easily observable properties for both scales. It can be easily seen that M is a Vietoris-Rips system if and only if

ν_{M} \leq 1

(in particular

ν_{M_{ν_{M}}} = 1

); similarly, M is a Čech system if and only if

μ_{M} \leq 1

Note that for a given d-disk system

M = {D_{1}, D_{2}, \dots, D_{m}}

the Vietoris-Rips scale is

ν_{M} = max_{i < j} {∥ c_{i} - c_{j} ∥ / (r_{i} + r_{j})}

where

c_{i}

and

r_{i}

are the center and radii of

D_{i}

. An additional observation is that, in cases where the disk system consists of either one or two disks, the Vietoris-Rips scale coincides with the Čech scale. It is evident that every Čech system is also a Vietoris-Rips system; however, the reverse assertion, in general, is not true.

Conversely, if the d-disk system contains at least three disks, determining the Čech scale becomes more complex. In the context of Čech scale, the following remark is important and play a key role in implementation (see [3] for details).

Remark 1.

μ_{M}

is the Čech scale for M, then the

μ_{M}

-rescaled system

M_{μ_{M}}

, has only one point in the intersection

⋂_{D_{i} \in M} D_{i} (c_{i}; μ_{M} r_{i})

As we have mentioned, a Čech system is also a Vietoris-Rips system, but the converse is not true. What we can affirm is that if a system is a Vietoris-Rips system, then the system rescaled by the factor

\sqrt{2 d / (d + 1)}

is also a Čech system. This is established by the following lemma, the proof of which can be found in [3].

Lemma 5.

Let

M = {D_{i} (c_{i}; r_{i})}

be a d-disk system in euclidean space

R^{d}

. If

D_{i} (c_{i}; r_{i}) \cap D_{j} (c_{j}; r_{j}) \neq \emptyset

for every pair of disks in M, then

⋂_{D_{i} \in M} D_{i} (c_{i}; \sqrt{2 d / (d + 1)} r_{i}) \neq \emptyset .

One of the implications of the previous result is that, for any given disk system M, we can bound the Čech scale using the Vietoris-Rips scale

ν_{M}

. This is stated by the following corollary.

Corollary 1.

If M is an arbitrary d-disk system and

ν_{M}

is its Vietoris-Rips scale, then its Čech scale satisfies

ν_{M} \leq μ_{M} \leq \sqrt{2 d / (d + 1)} ν_{M}

. Therefore, for every d-disk system M, the rescaled disk system

M_{\sqrt{2 d / (d + 1)} ν_{M}}

is always a Čech system. In particular, if

\sqrt{2 d / (d + 1)} ν_{M} \leq 1

then

M_{ν_{M}}

is a Čech system.

3.1. Algorithm for determining Čech system.

In the previous section, we have determined the

e_{q}

-poles for the intersection of any number of disks in

R^{d}

. If any of these poles is in all disks of the disk system, it indicates that the system conforms to the criteria of a Čech system. It is important to recognize that this result streamlines our calculation process, focusing on specific points to establish whether the system exhibits a non-empty intersection.

Given a system of m disks in

R^{d}

where

m > d

, it is enough to verify if every subsystem of

d + 1

disks qualifies as a Čech system to conclude that the entire system of disks has a non-empty intersection. This assertion is supported by the Helly’s Theorem.

Now, we introduce an algorithm that determines whether a disk system qualifies as a Čech system. In simpler terms, if the disk system exhibits a non-empty intersection, the algorithm outputs "TRUE"; otherwise, it outputs "FALSE". The algorithm operates by seeking poles within the intersections of the disk boundaries, which, as we have observed, correspond to i-spheres. It initiates the search for poles within individual disks and then progresses to the pairwise intersections of the disk boundaries (

(d - 2) - s p h e r e

), continuing the process iteratively. If a pole is found within the remaining disks, the system is classified as a Čech system.

Theorem 1.

Let

M = {D_{1}, \dots, D_{m}}

be a d-disk system. Then, M is a Čech system if and only if Cech.system $(M) =$ TRUE.

Proof.

If Cech.system $(M) =$ TRUE, then the Cech.system algorithm (Algorithm 1) found a pole contained in the intersection

\cap_{j = 1}^{m} D_{j}

, therefore

\cap_{j = 1}^{m} D_{j} \neq \emptyset

and it follows that M is a Čech system.

On the other hand, if

⋂_{j = 1}^{m} D_{j} \neq \emptyset

, let p be a point in

⋂_{j = 1}^{m} D_{j}

satisfying

π_{1} (p) \leq π_{1} (x)

for every x in

⋂_{j = 1}^{m} D_{j}

. By Lemma 2, it follows that p belongs to an i-sphere and must be an

e_{1}

-south pole. Therefore, by the exhaustive search of Algorithm 1 across all poles, its output is Cech.system $(M) =$ TRUE. □

Algorithm 1: Cech.system

3.2. Algorithm to compute the Čech scale

Finding the minimum parameter for which the rescaled system of disks has a non-empty intersection is significant because it helps identify a critical threshold at which the disks come into contact. This parameter, known as the Čech scale, provides valuable information about the proximity or overlap of the disks, which can be crucial in various applications such as collision detection in computer graphics, spatial packing problems, and modeling physical phenomena. In this section, we introduce an algorithm to compute an approximation of the Čech scale for a system of m disks in

R^{d}

Algorithm 2: Cech.scale

The given code presents an algorithm to compute an approximation of the Čech scale of a disk system in Euclidean space using Algorithm 1 and a precision parameter

η > 0

. It initializes the scale factor

λ

to the Rips scale

ν_{M}

. If this scale satisfies Cech.system

(M_{ν_{M}}) =

TRUE, it indicates that the Čech scale has been found. Otherwise, we initiate a cycle in which we compute Cech.system of the system rescaled by a factor

λ

. The Čech scale is known to fall between the Rips scale and the value

\sqrt{2 d / (d + 1)} ν_{M}

(Generalized Vietoris-Rips, Corollary 1). To approximate the Čech scale, we employ the bisection method as long as the interval enclosing the Čech scale has a length greater than

η

. Finally, the algorithm returns an approximation of the Čech scale.

Utilizing the previously described algorithm, we can construct the filtered generalized Čech complex for a disk system M. Let

C (M)

denote the set of Čech subsystems, and

C_{λ} (M)

the set of Čech subsystems for the rescaled disk system

M_{λ}

. The Čech filtration of the M system forms a maximal chain of Čech complexes

C_{*} (M) : C_{0} (M) ⊊ C_{λ_{1}} (M) ⊊ C_{λ_{2}} (M) ⊊ . . . ⊊ C_{μ_{M}} (M),

where each

λ_{i}

represents the Čech scale of the system

M_{λ_{i}}

. Since the Čech scale of a disk system indicates the factor by which we must rescale the system to make it Čech, defining a level of the filtration,

C_{λ} (M)

, simply requires determining the Čech scale of the system

M_{λ}

4. Minimal Axis-Aligned Bounding Box.

In this section, we introduce the concept of the minimal axis-aligned bounding box (AABB) for the intersection of d-disks and present methods for its computation. The AABB provides a simplified representation of the disk intersection, making it easier to obtain valuable information about the disks. This information could be useful for computing the Čech scale of a disk system.

Definition 4.

Let M be a disk system in

R^{d}

. The minimal axis-aligned bounding box of M, denoted as

A A B B (M)

, is defined as the smallest axis-aligned bounding box that contains the intersection

D = \cap_{D_{i} \in M} D_{i}

, given by

A A B B (M) : = ⋂_{D \subset \tilde{B}} \tilde{B}

where

\tilde{B}

ranges over all axis-aligned bounding boxes that contain D.

Note that the AABB can be expressed as

A A B B (M) = \prod_{k = 1}^{d} [inf π_{k} (\partial D), sup π_{k} (\partial D)]

where

π_{k} : R^{d} ⟶ R

is the canonical projection onto the k-th factor, and

\partial D

denotes the boundary of D. In other words, the AABB of a disk system M is given by the Cartesian product of intervals, where each interval is determined by the minimum and maximum values of the corresponding projection of the disk boundaries.

4.1. Minimal axis-aligned bounding box for two disks

Let’s consider the situation when the disk system M is composed of two disks

D_{1}

and

D_{2}

R^{d}

. If

D_{1} \cap D_{2} \neq \emptyset

and

D_{1} \neq D_{2}

, the subset

\partial D_{1} \cap \partial D_{2}

can take one of three forms: an empty set, a single common point (when the disks are tangent), or a

(d - 2)

-dimensional sphere. In the last case, we denote the

(d - 2)

-dimensional sphere or the

(d - 1)

-sphere

\partial D_{1} \cap \partial D_{2}

S_{1, 2}

. To calculate

inf π_{i} (\partial (D_{1} \cap D_{2}))

and

sup π_{i} (\partial (D_{1} \cap D_{2}))

for each

i \in {1, 2, . . ., d}

, we can use:

\begin{matrix} inf π_{i} (\partial (D_{1} \cap D_{2})) = \{\begin{matrix} π_{i} (c_{1} - r_{1} e_{i}) & if c_{1} - r_{1} e_{i} \in D_{2}, \\ π_{i} (c_{2} - r_{2} e_{i}) & if c_{2} - r_{2} e_{i} \in D_{1}, \\ inf (π_{i} (\partial D_{1} \cap \partial D_{2})) & otherwise, \end{matrix} \end{matrix}

(6)

\begin{matrix} sup π_{i} (\partial (D_{1} \cap D_{2})) = \{\begin{matrix} π_{i} (c_{1} + r_{1} e_{i}) & if c_{1} + r_{1} e_{i} \in D_{2}, \\ π_{i} (c_{2} + r_{2} e_{i}) & if c_{2} + r_{2} e_{i} \in D_{1}, \\ sup (π_{i} (\partial D_{1} \cap \partial D_{2})) & otherwise . \end{matrix} \end{matrix}

Indeed, by Lemma 2, the extremes of the AABB are the projections of certain poles, either from the

(d - 1)

-sphere or some d-sphere. The d-spheres represent the boundaries of each disk, with poles given by

c_{j} \pm r_{j} e_{i}

, and the

(d - 1)

-sphere is

S_{1, 2} = \partial D_{1} \cap \partial D_{2}

, whose poles are computed using Lemma 4. It is worth noting that there are no further options for

(d - m + 1)

-spheres in the case of a two-disk system.

To simplify the notation, we will use

B_{i, j}

to denote the AABB of the intersection of disks

D_{i}

and

D_{j}

. Figure 2 illustrates the AABB of the intersection of two disks in the plane.

According to (6) and Lemma 3, we have a method to calculate the axis-aligned boundary box (AABB) for systems of two disks.

Knowing how to compute the AABB of two disks is not sufficient to determine the AABB for a disk system with more than two disks in

R^{d}

. In the following examples, we demonstrate that the AABB of a disk system is not simply the intersection of all AABBs of two disks in

R^{3}

Example 1. Let

M = {D_{1} ((4, 1, 0); \sqrt{2}), D_{2} ((4, - 1, 0); \sqrt{2}), D_{3} ((0, 0, 0); 3)} \subset R^{3}

be a Vietoris-Rips system in

R^{3}

with the following projection onto the

x y

-plane:

By computing the boxes

B_{i, j}

for all

i < j

, we obtain:

\begin{matrix} B_{1, 2} & = [3, 5] \times [- \sqrt{2}, \sqrt{2}] \times [- 1, 1] \\ B_{1, 3} & = [4 - \sqrt{2}, 3] \times [0, 1.41] \times [- 0.72, 0.72] \\ B_{2, 3} & = [4 - \sqrt{2}, 3] \times [- 1.4, 0] \times [- 0.72, 0.72] \end{matrix}

Therefore,

\cap_{i < j} B_{i, j} = [3, 3] \times [0, 0] \times [- 0.72, 0.72]

. However, the disk system intersects at the point

P = (3, 0, 0)

, which means

A A B B (M) = {P}

. In other words,

A A B B (M)

is not equal to

\cap_{i < j} B_{i, j}

We know that if

D \neq \emptyset

, then

\cap_{i < j} B_{i, j} \neq \emptyset

. However, the converse is not always true. The following example illustrates this fact.

Example 2. Let

N = {D_{1}, D_{2}, D_{3} ((0, 1, 0); \sqrt{10}), D_{4} ((3, 0, 1); 0.9)} \subset R^{3}

be a Vietoris-Rips system. The projection of disks

D_{1}, D_{2},

and

D_{3}

onto the

x y

-plane is illustrated in Figure 4.

We will now compute the intersections

B_{i, j}

for different pairs of disks:

\begin{matrix} B_{1, 2} & = [3, 5] \times [- \sqrt{2}, \sqrt{2}] \times [- 1, 1] \\ B_{1, 3} & = [4 - \sqrt{2}, \sqrt{10}] \times [0, 2] \times [- 1, 1] \\ B_{1, 4} & = [2.6, 3.9] \times [- 0.4, 0.9] \times [0.100007, 1.3] \\ B_{2, 3} & = [2.6, 3] \times [- 0.8, 0] \times [- 0.44, 4.47] \\ B_{2, 4} & = [2.6, 3.9] \times [- 0.9, 0.4] \times [0.100007, 1.3] \\ B_{3, 4} & = [2.1, 3.11] \times [- 0.73, 0.9] \times [0.1, 1.7] \end{matrix}

The intersection of all pairwise intersections,

\cap_{i < j} B_{i, j}

, is given by

[3, 3] \times [0, 0] \times [0.1, 1.7]

. However, the intersection of disks

D_{1}, D_{2},

and

D_{3}

is a single point P, which is not contained in

D_{4}

(by construction). Therefore, D is an empty set, but

\cap_{i < j} B_{i, j}

is not.

These examples clearly illustrate that when dealing with the AABB of three disks or more, knowing the AABB for pairs of disks is insufficient. In Example Section 4.1, we observe that the intersection of

A A B B ({D_{i}, D_{j}})

is not equal to the AABB of the disk system M. Similarly, in Example Section 4.1, we find that the intersection of three disks is empty, yet the intersection of

A A B B ({D_{i}, D_{j}})

contains points.

Therefore, the next crucial step is to determine how to calculate the AABB of a disk system consisting of more than two disks in

R^{d}

4.2. Minimal axis-aligned bounding box for more than two disks

Given a system M consisting of m disks in

R^{d}

, we can compute the

e_{q}

-poles for any subcollection of disks. Using these

e_{q}

-poles, we can determine the axis-aligned bounding box (AABB) of M.

m \leq d

, we calculate the

e_{q}

-poles of the

(d - m + 1)

-sphere

\cap \partial D_{i}

, and with these poles, we define the AABB of M by taking

inf π_{q} \partial D = π_{q} (p)

, where p is the

e_{q}

-south pole of the

(d - m + 1)

-sphere (similarly for

sup π_{q} \partial D

In the case where

m = d + 1

, we consider

A A B B (M (1)), . . ., A A B B (M (m))

as a collection of minimal axis-aligned boxes for the disk system

M (i) = M - {D_{i}}

R^{d}

. If the intersection of any

d + 1

of these sets is nonempty, then the intersection of the entire collection gives us the minimal axis-aligned box of the disk system M. This can be expressed as

A A B B (M) = \cap A A B B (M (i))

. The next theorem confirms this finding.

Theorem 2

(Helly’s theorem for minimal axis-aligned boxes). Let M be a Rips system with

d + 1

disks in

R^{d}

. Then, the minimal axis-aligned bounding box for the intersection set

D = \cap_{j = 1}^{d + 1} D_{j}

satisfies:

A A B B (M) = ⋂_{j = 1}^{d + 1} A A B B (M (j))

where

M (j) = M - {D_{j}}

Proof.

We know that

A A B B (M) \subseteq ⋂_{j = 1}^{d + 1} A A B B (M (j))

. Now, our objective is to establish the reverse inclusion, that is,

⋂_{j = 1}^{d + 1} A A B B (M (j)) \subseteq A A B B (M)

. In order to derive a contradiction, suppose that the reverse inclusion is not true. By definition, we have:

A A B B (M) = \prod_{i = 1}^{d} [inf π_{i} (\partial D), sup π_{i} (\partial D)]

and

⋂_{j = 1}^{d + 1} A A B B (M (j)) = \prod_{i = 1}^{d} [max_{k} {inf π_{i} (\partial \cap_{j \neq k} D_{j})}, min_{k} {sup π_{i} (\partial \cap_{j \neq k} D_{j})}]

Without loss of generality, let’s assume that:

π_{1} (p) > max_{k} {\{inf π_{1} (\cap_{j \neq k} \partial D_{j})\}}_{k = 1}^{d + 1}

where

p \in D

satisfies

π_{1} (p) = inf π_{1} (\partial D)

Let

q_{j}

be the point in

\cap_{k \neq j} D_{k}

such that

π_{1} (q_{j}) = inf π_{1} (\cap_{k \neq j} D_{k})

for each

j = 1, 2, . . ., d + 1

. Note that

q_{j} \notin D_{j}

because

q_{j}

is not in D. Now, let

γ_{j}

be the line segment that connects

q_{j}

and p.

Choose

ϵ > 0

small enough such that the hyperplane

P : x_{1} = π_{1} (p) - ϵ

does not contain any

q_{j}

, and

π_{1} (q_{j}) < π_{1} (p) - ϵ

for all

j = 1, 2, . . ., d + 1

(such hyperplane P exists because

π_{1} (p) > π_{1} (q_{j})

for all

j = 1, . . ., d + 1

). Since

q_{j}

is in

\cap_{k \neq j} D_{k}

, the hyperplane P intersects every disk in

M (j)

for each j, and

γ_{j} \subset \cap_{k \neq j} D_{k}

intersects P at a point that is in

\cap_{k \neq j} D_{k}

(see Figure 5). Furthermore,

D_{k} \cap P

is a

(d - 1)

-dimensional disk. Therefore, we have a collection

D = {D_{j} \cap P | j = 1, 2, . . ., d + 1}

d + 1

disks, each of dimension

d - 1

, such that every subset A of

D

consisting of d disks has the non-empty intersection property. By Helly’s Theorem, the intersection of all

(d - 1)

-disks in

D

is not empty. Therefore, there exists a point

q \in D

with

π_{1} (q) < π_{1} (p)

. However, this contradicts the fact that

p \in D

is such that

π_{1} (p) = inf π_{1} (\partial D)

Therefore, we conclude that

⋂_{j = 1}^{d + 1} A A B B (M (j)) = A A B B (M)

□

Lemma 6.

Let M be a Vietoris-Rips system of

d + 1

disks in

R^{d}

. If

D = \cap_{i = 1}^{d + 1} D_{i} = \emptyset

and

A A B B (M (j)) \neq \emptyset

for each

j = 1, . . ., d + 1

, then

\cap_{i = 1}^{d + 1} A A B B (M (i))

consists only of intervals of the form

[a, b]

with

a > b

(inverted intervals).

Proof.

Suppose that

\cap_{i = 1}^{d + 1} A A B B (M (i))

contains an interval. Without loss of generality, let’s assume it is the interval

[inf π_{1} (\partial D (k)), sup π_{1} (\partial D (l))]

for some

k, l \in {1, 2, . . ., d + 1}

where

D (k) = \cap_{i \neq k} D_{i}

. By definition of

\cap_{i = 1}^{d + 1} A A B B (M (i))

, we have

inf π_{1} (\partial D (k)) \geq inf π_{1} (\partial D (i))

for all

i \neq k

and

sup π_{1} (\partial D (l)) \leq sup π_{1} (\partial D (j))

for all

j \neq l

. Now, consider a hyperplane

P : x_{1} = p

with

inf π_{1} (\partial D (k)) \leq p \leq sup π_{1} (\partial D (k))

. The hyperplane P intersects

D (j)

for every

j = 1, . . ., d + 1

(since P cuts through all the boxes

A A B B (M (j))

). We also know that

P \cap D_{i}

is a

d - 1

-dimensional disk for each i.

Therefore, we have a collection of d disks in a

d - 1

dimensional space, and by Helly’s Theorem, this collection must have a non-empty intersection. This implies that

D \neq \emptyset

, which contradicts our assumption that D is empty. Hence, all intervals in

\cap_{i = 1}^{d + 1} A A B B (M (i))

must be inverted intervals of the form

[a, b]

with

a > b

. □

As an example, we provide the lower bounds of the AABB for a system of three disks in

R^{d}

, this is, we compute

inf π_{i} (\partial D)

for each

i = 1, . . ., d

(analogous

sup π_{i} \partial D

) with

D = \cap_{j = 1}^{3} D_{j}

. Let

M = {D_{1}, D_{2}, D_{3}}

be a Rips system in

R^{d}

, then, for each

i = 1, . . ., d

, the computation of the AABB for the disk system M is given by:

\begin{matrix} inf π_{i} (\partial D) = \{\begin{matrix} π_{i} (c_{k} - r_{k} e_{i}) & if c_{k} - r_{k} e_{i} \in D \\ max_{j < k} {inf π_{i} (\partial D_{j} \cap \partial D_{k})} & if (*) \\ inf π_{i} (\cap_{j = 1}^{3} \partial D_{j}) & otherwise \end{matrix} \end{matrix}

where (*) denotes the case where

q \in D

with

q \in \partial D_{j} \cap \partial D_{k}

such that

π_{i} (q) = max_{j < k} {inf π_{i} (\partial D_{j} \cap \partial D_{k})} .

The preceding calculations are a result of Lemma 2. It is known that the extremes of the AABB lie in the projections of specific poles, either from a d-sphere,

(d - 1)

-sphere, or the

(d - 2)

-sphere.

Given a disk system M in

R^{d}

, we can compute the minimal axis-aligned bounding box (AABB) for the intersection of all disks in M. If the AABB is a point, then it represents the intersection of the disks. This property allows us to identify when the AABB is a point. If the AABB of M is not a point, we can rescale M by a scale factor

λ

such that the AABB of

M_{λ}

becomes a point. The value of

λ

is referred to as the Čech scale of the system M.

4.3. Minimal axis-aligned bounding box Algorithm

Now, we present an algorithm to calculate the minimal axis-aligned bounding box (AABB) of a system of m disks in

R^{d}

. As previously explained, the AABB’s extremes are defined by projecting the poles of specific i-spheres. Thus, in computing the AABB, we will determine the poles for each i-sphere within the disk system.

The algorithm starts by initializing the set of poles, P, as an empty set. In each iteration of the first loop, we determine the spheres formed by the intersection of the boundaries of the disk subcollections in the system M (for this, we require the center and radius, which are computed in Section 2.2). Next, we identify the poles of each sphere and if any of them are present in all the disks of M, we add them to the set P. If P is not empty, we proceed to calculate the extremes for each dimension of the AABB using the set of north poles for the upper bounds and the set of south poles for the lower bounds. In this case, the output is the product of the intervals defined by the computed extremes. If P is empty, it indicates that there is no intersection in the disk system.

Algorithm 3: AABB.minimal

Acknowledgments

C.G.E.P acknowledges CONACYT for the financial support provided through a National Fellowship (CVU-638165).

References

Bell, G.; Lawson, A.; Martin, J.; Rudzinski, J.; Smyth, C. Weighted persistent homology. Involve, a Journal of Mathematics 2019, 12, 823–837. [Google Scholar] [CrossRef]
Le, N.K.; Martins, P.; Decreusefond, L.; Vergne, A. Construction of the generalized Cech complex, 2015, [arXiv:cs.NI/1409.8225]. arXiv:cs.NI/1409.8225].
Espinoza, J.F.; Hernández-Amador, R.; Hernández-Hernández, H.A.; Ramonetti-Valencia, B. A Numerical Approach for the Filtered Generalized Čech Complex. Algorithms 2019, 13, 11. [Google Scholar] [CrossRef]
Mahovsky, J.; Wyvill, B. Fast Ray-Axis Aligned Bounding Box Overlap Tests with Plucker Coordinates. Journal of Graphics Tools 2004, 9, 35–46. [Google Scholar] [CrossRef]
Cai, P.; Cai, Y.; Chandrasekaran, I.; Zheng, J. Collision detection using axis aligned bounding boxes. Simulation, Serious Games and Their Applications 2013, pp. 1–14.
Maioli, D.S.; Lavor, C.; Gonçalves, D.S. A note on computing the intersection of spheres in Rⁿ. The ANZIAM Journal 2017, 59, 271–279. [Google Scholar] [CrossRef]

Figure 1. Visualization of the subspaces L and

M + {c}

Figure 1. Visualization of the subspaces L and

M + {c}

Figure 2. AABB of two disks.

Figure 3. Disk system M projected onto the

x y

-plane.

Figure 3. Disk system M projected onto the

x y

-plane.

Figure 4. Disk system N projected onto the

x y

-plane.

Figure 4. Disk system N projected onto the

x y

-plane.

Figure 5.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.

MDPI Initiatives

Important Links

Choose an area of interest and we will send you notifications of new preprints at your preferred frequency.

Disclaimer

Intersection Properties of Finite Disk Collections

Abstract

1. Introduction

2. Intersection properties of sphere systems

2.1. Sphere systems with two spheres

2.2. Sphere systems with more than two spheres

3. Vietoris-Rips and Čech systems

3.1. Algorithm for determining Čech system.

3.2. Algorithm to compute the Čech scale

4. Minimal Axis-Aligned Bounding Box.

4.1. Minimal axis-aligned bounding box for two disks

4.2. Minimal axis-aligned bounding box for more than two disks

4.3. Minimal axis-aligned bounding box Algorithm

Acknowledgments

References

MDPI Initiatives

Important Links

Subscribe