Simultaneous inner and outer approximation of shapes

Rudolf Fleischer, Kurt Mehlhorn, Günter Rote, Emo Welzl, Chee Yap

Research output: Contribution to journalArticle

Abstract

For compact Euclidean bodies P, Q, we define λ(P, Q) to be the smallest ratio r/s where r > 0, s > 0 satisfy {Mathematical expression}. Here sQ denotes a scaling of Q by the factor s, and Q′, Q″ are some translates of Q. This function λ gives us a new distance function between bodies which, unlike previously studied measures, is invariant under affine transformations. If homothetic bodies are identified, the logarithm of this function is a metric. (Two bodies are homothetic if one can be obtained from the other by scaling and translation.) For integer k ≥ 3, define λ(k) to be the minimum value such that for each convex polygon P there exists a convex k-gon Q with λ(P, Q) ≤ λ(k). Among other results, we prove that 2.118 ... <-λ(3) ≤ 2.25 and λ(k) = 1 + Θ(k -2). We give an O(n 2 log2 n)-time algorithm which, for any input convex n-gon P, finds a triangle T that minimizes λ(T, P) among triangles. However, in linear time we can find a triangle t with λ(t, P)<-2.25. Our study is motivated by the attempt to reduce the complexity of the polygon containment problem, and also the motion-planning problem. In each case we describe algorithms which run faster when certain implicit slackness parameters of the input are bounded away from 1. These algorithms illustrate a new algorithmic paradigm in computational geometry for coping with complexity.

Original languageEnglish (US)
Pages (from-to)365-389
Number of pages25
JournalAlgorithmica (New York)
Volume8
Issue number1-6
DOIs
StatePublished - Dec 1992

Fingerprint

Outer Approximation
Triangle
Scaling
n-gon
Computational geometry
Convex polygon
Motion Planning
Computational Geometry
Distance Function
Motion planning
Logarithm
Polygon
Affine transformation
Linear Time
Euclidean
Paradigm
Denote
Minimise
Metric
Integer

Keywords

  • Algorithmic paradigms
  • Banach-Mazur metric
  • Computational geometry
  • Implicit complexity parameters
  • Polygonal approximation
  • Shape approximation

ASJC Scopus subject areas

  • Applied Mathematics
  • Safety, Risk, Reliability and Quality
  • Software
  • Computer Graphics and Computer-Aided Design
  • Computer Science Applications
  • Computer Science(all)

Cite this

Simultaneous inner and outer approximation of shapes. / Fleischer, Rudolf; Mehlhorn, Kurt; Rote, Günter; Welzl, Emo; Yap, Chee.

In: Algorithmica (New York), Vol. 8, No. 1-6, 12.1992, p. 365-389.

Research output: Contribution to journalArticle

Fleischer, R, Mehlhorn, K, Rote, G, Welzl, E & Yap, C 1992, 'Simultaneous inner and outer approximation of shapes', Algorithmica (New York), vol. 8, no. 1-6, pp. 365-389. https://doi.org/10.1007/BF01758852
Fleischer, Rudolf ; Mehlhorn, Kurt ; Rote, Günter ; Welzl, Emo ; Yap, Chee. / Simultaneous inner and outer approximation of shapes. In: Algorithmica (New York). 1992 ; Vol. 8, No. 1-6. pp. 365-389.
@article{b3237c63cb9d41b3b071b5ee732418a0,
title = "Simultaneous inner and outer approximation of shapes",
abstract = "For compact Euclidean bodies P, Q, we define λ(P, Q) to be the smallest ratio r/s where r > 0, s > 0 satisfy {Mathematical expression}. Here sQ denotes a scaling of Q by the factor s, and Q′, Q″ are some translates of Q. This function λ gives us a new distance function between bodies which, unlike previously studied measures, is invariant under affine transformations. If homothetic bodies are identified, the logarithm of this function is a metric. (Two bodies are homothetic if one can be obtained from the other by scaling and translation.) For integer k ≥ 3, define λ(k) to be the minimum value such that for each convex polygon P there exists a convex k-gon Q with λ(P, Q) ≤ λ(k). Among other results, we prove that 2.118 ... <-λ(3) ≤ 2.25 and λ(k) = 1 + Θ(k -2). We give an O(n 2 log2 n)-time algorithm which, for any input convex n-gon P, finds a triangle T that minimizes λ(T, P) among triangles. However, in linear time we can find a triangle t with λ(t, P)<-2.25. Our study is motivated by the attempt to reduce the complexity of the polygon containment problem, and also the motion-planning problem. In each case we describe algorithms which run faster when certain implicit slackness parameters of the input are bounded away from 1. These algorithms illustrate a new algorithmic paradigm in computational geometry for coping with complexity.",
keywords = "Algorithmic paradigms, Banach-Mazur metric, Computational geometry, Implicit complexity parameters, Polygonal approximation, Shape approximation",
author = "Rudolf Fleischer and Kurt Mehlhorn and G{\"u}nter Rote and Emo Welzl and Chee Yap",
year = "1992",
month = "12",
doi = "10.1007/BF01758852",
language = "English (US)",
volume = "8",
pages = "365--389",
journal = "Algorithmica",
issn = "0178-4617",
publisher = "Springer New York",
number = "1-6",

}

TY - JOUR

T1 - Simultaneous inner and outer approximation of shapes

AU - Fleischer, Rudolf

AU - Mehlhorn, Kurt

AU - Rote, Günter

AU - Welzl, Emo

AU - Yap, Chee

PY - 1992/12

Y1 - 1992/12

N2 - For compact Euclidean bodies P, Q, we define λ(P, Q) to be the smallest ratio r/s where r > 0, s > 0 satisfy {Mathematical expression}. Here sQ denotes a scaling of Q by the factor s, and Q′, Q″ are some translates of Q. This function λ gives us a new distance function between bodies which, unlike previously studied measures, is invariant under affine transformations. If homothetic bodies are identified, the logarithm of this function is a metric. (Two bodies are homothetic if one can be obtained from the other by scaling and translation.) For integer k ≥ 3, define λ(k) to be the minimum value such that for each convex polygon P there exists a convex k-gon Q with λ(P, Q) ≤ λ(k). Among other results, we prove that 2.118 ... <-λ(3) ≤ 2.25 and λ(k) = 1 + Θ(k -2). We give an O(n 2 log2 n)-time algorithm which, for any input convex n-gon P, finds a triangle T that minimizes λ(T, P) among triangles. However, in linear time we can find a triangle t with λ(t, P)<-2.25. Our study is motivated by the attempt to reduce the complexity of the polygon containment problem, and also the motion-planning problem. In each case we describe algorithms which run faster when certain implicit slackness parameters of the input are bounded away from 1. These algorithms illustrate a new algorithmic paradigm in computational geometry for coping with complexity.

AB - For compact Euclidean bodies P, Q, we define λ(P, Q) to be the smallest ratio r/s where r > 0, s > 0 satisfy {Mathematical expression}. Here sQ denotes a scaling of Q by the factor s, and Q′, Q″ are some translates of Q. This function λ gives us a new distance function between bodies which, unlike previously studied measures, is invariant under affine transformations. If homothetic bodies are identified, the logarithm of this function is a metric. (Two bodies are homothetic if one can be obtained from the other by scaling and translation.) For integer k ≥ 3, define λ(k) to be the minimum value such that for each convex polygon P there exists a convex k-gon Q with λ(P, Q) ≤ λ(k). Among other results, we prove that 2.118 ... <-λ(3) ≤ 2.25 and λ(k) = 1 + Θ(k -2). We give an O(n 2 log2 n)-time algorithm which, for any input convex n-gon P, finds a triangle T that minimizes λ(T, P) among triangles. However, in linear time we can find a triangle t with λ(t, P)<-2.25. Our study is motivated by the attempt to reduce the complexity of the polygon containment problem, and also the motion-planning problem. In each case we describe algorithms which run faster when certain implicit slackness parameters of the input are bounded away from 1. These algorithms illustrate a new algorithmic paradigm in computational geometry for coping with complexity.

KW - Algorithmic paradigms

KW - Banach-Mazur metric

KW - Computational geometry

KW - Implicit complexity parameters

KW - Polygonal approximation

KW - Shape approximation

UR - http://www.scopus.com/inward/record.url?scp=52449148577&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=52449148577&partnerID=8YFLogxK

U2 - 10.1007/BF01758852

DO - 10.1007/BF01758852

M3 - Article

AN - SCOPUS:0000099132

VL - 8

SP - 365

EP - 389

JO - Algorithmica

JF - Algorithmica

SN - 0178-4617

IS - 1-6

ER -