Helly-Type Theorems for Approximate Covering

Let <i>F</i> ∪ {<i>U</i>} be a collection of convex sets in R<sup>d</sup> such that <i>F</i> covers <i>U</i>. We show that if the elements of <i>F</i> and <i>U</i> have comparable size, in the sense that each contains a ball of radius <i>r</i> and is contained in a ball of radius <i>R</i> for some fixed <i>r</i> and <i>R</i>, then for any ε > 0 there exists <i>H</i><sub>ε</sub> ⊂ <i>F</i>, whose size |<i>H</i><sub>ε</sub>| is polynomial in 1/ε and independent of |<i>F</i>|, that covers <i>U</i> except for a volume of at most ε. The size of the smallest such subset depends on the geometry of the elements of <i>F</i>; specifically, we prove that it is <i>O</i>(1/ε) when <i>F</i> consists of axis-parallel unit squares in the plane and <i>Õ</i>(ε<sup>1--<i>d</i>/2</sup>) when <i>F</i> consists of unit balls in R<sup>d</sup> (here, <i>Õ</i>(<i>n</i>) means <i>O</i>(<i>n</i> log <i>n</i>) for some constant), and that these bounds are, in the worst-case, tight up to the logarithmic factors. We extend these results to surface-to-surface visibility in 3 dimensions: if a collection <i>F</i> of disjoint unit balls occludes visibility between two balls then a subset of <i>F</i> of size <i>Õ</i>(ε<sup>--7/2</sup>) blocks visibility along all but a set of lines of measure ε. Finally, for each of the above situations we give an algorithm that takes <i>F</i> and <i>U</i> as input and outputs in time <i>O</i>(|<i>F</i>|*|<i>H</i><sub>ε</sub>|) either a point in <i>U</i> not covered by <i>F</i> or a subset <i>H</i><sub>ε</sub> covering <i>U</i> up to a measure ε, with |<i>H</i><sub>ε</sub>| satisfying the above bounds.

[1]  L. Santaló Integral geometry and geometric probability , 1976 .

[2]  Jörg M. Wills,et al.  Handbook of Convex Geometry , 1993 .

[3]  Nina Amenta,et al.  Helly theorems and generalized linear programming , 1993, SCG '93.

[4]  Raimund Seidel,et al.  Small-dimensional linear programming and convex hulls made easy , 1991, Discret. Comput. Geom..

[5]  Micha Sharir,et al.  A Combinatorial Bound for Linear Programming and Related Problems , 1992, STACS.

[6]  V. Klee,et al.  Helly's theorem and its relatives , 1963 .

[7]  Michael F. Cohen,et al.  Radiosity and realistic image synthesis , 1993 .

[8]  J. Pach,et al.  Helly"s theorem with volumes , 1984 .

[9]  Claude Puech,et al.  Radiosity and global illumination , 1994 .

[10]  Xavier Goaoc,et al.  Helly-Type Theorems for Approximate Covering , 2009, Discret. Comput. Geom..

[11]  M. Sharir Combinatorial Geometry with Algorithmic Applications The Alcala Lectures János Pach , 2008 .

[12]  Marco Pellegrini,et al.  Monte carlo approximation of form factors with error bounded a priori , 1995, SCG '95.

[13]  J. Eckhoff Helly, Radon, and Carathéodory Type Theorems , 1993 .

[14]  Michael T. Goodrich,et al.  Almost optimal set covers in finite VC-dimension , 1995, Discret. Comput. Geom..

[15]  J. Pach,et al.  Quantitative Helly-type theorems , 1982 .

[16]  Micha Sharir,et al.  Combinatorial Geometry and Its Algorithmic Applications , 2008 .

[17]  M. Birkner,et al.  Blow-up of semilinear PDE's at the critical dimension. A probabilistic approach , 2002 .

[18]  Carsten Lund,et al.  On the hardness of approximating minimization problems , 1993, STOC.

[19]  Micha Sharir,et al.  Combinatorial Geometry with Algorithmic Applications , 2006 .

[20]  Sunil Arya,et al.  The effect of corners on the complexity of approximate range searching , 2006, SCG '06.

[21]  Rephael Wenger,et al.  Helly-Type Theorems and Geometric Transversals , 2004, Handbook of Discrete and Computational Geometry, 2nd Ed..

[22]  E. Helly Über Mengen konvexer Körper mit gemeinschaftlichen Punkte. , 1923 .

[23]  Kenneth L. Clarkson,et al.  Improved Approximation Algorithms for Geometric Set Cover , 2007, Discret. Comput. Geom..

[24]  Emo Welzl,et al.  Linear Programming - Randomization and Abstract Frameworks , 1996, STACS.

[25]  Nina Amenta,et al.  Helly-type theorems and Generalized Linear Programming , 1994, Discret. Comput. Geom..

[26]  Mordecai J. Golin How Many Maxima Can There Be? , 1992, Comput. Geom..

[27]  Raymond E. Miller,et al.  Complexity of Computer Computations , 1972 .