relative neighborhood graph

File:Relative neighborhood graph.svg

In computational geometry, the relative neighborhood graph (RNG) is an undirected graph defined on a set of points in the Euclidean plane by connecting two points p and q by an edge whenever there does not exist a third point r that is closer to both p and q than they are to each other. This graph was proposed by Godfried Toussaint in 1980 as a way of defining a structure from a set of points that would match human perceptions of the shape of the set.{{citation

| last = Toussaint | first = G. T. | author-link = Godfried Toussaint

| doi = 10.1016/0031-3203(80)90066-7

| journal = Pattern Recognition

| pages = 261–268

| title = The relative neighborhood graph of a finite planar set

| volume = 12

| year = 1980

| issue = 4}}.{{citation

| last1 = Jaromczyk | first1 = J.W.

| last2 = Toussaint | first2 = G.T.

| doi = 10.1109/5.163414

| issue = 9

| journal = Proceedings of the IEEE

| pages = 1502–1517

| title = Relative neighborhood graphs and their relatives

| volume = 80

| year = 1992}}.

Algorithms

{{harvtxt|Supowit|1983}} showed how to construct the relative neighborhood graph of n points in the plane efficiently in O(n\log n) time.{{citation

| last = Supowit | first = K. J.

| doi = 10.1145/2402.322386

| issue = 3

| journal = Journal of the ACM

| pages = 428–448

| title = The relative neighborhood graph, with an application to minimum spanning trees

| volume = 30

| year = 1983| doi-access = free

}}. It can be computed in O(n) expected time, for random set of points distributed uniformly in the unit square.{{citation

| last1 = Katajainen | first1 = Jyrki

| last2 = Nevalainen | first2 = Olli

| last3 = Teuhola | first3 = Jukka

| doi = 10.1016/0020-0190(87)90225-0

| issue = 2

| journal = Information Processing Letters

| pages = 77–86

| title = A linear expected-time algorithm for computing planar relative neighbourhood graphs

| volume = 25

| year = 1987}}. The relative neighborhood graph can be computed in linear time from the Delaunay triangulation of the point set.{{citation

| last1 = Jaromczyk | first1 = J. W.

| last2 = Kowaluk | first2 = M.

| contribution = A note on relative neighborhood graphs

| doi = 10.1145/41958.41983

| location = New York, NY, USA

| pages = 233–241

| publisher = ACM

| title = Proc. 3rd Symp. Computational Geometry

| year = 1987| isbn = 0-89791-231-4

}}.{{citation

| last = Lingas | first = A.

| doi = 10.1016/0925-7721(94)90018-3

| issue = 4

| journal = Computational Geometry

| pages = 199–208

| title = A linear-time construction of the relative neighborhood graph from the Delaunay triangulation

| volume = 4

| year = 1994| doi-access = free

}}.

Generalizations

Because it is defined only in terms of the distances between points, the relative neighborhood graph can be defined for point sets in any {{nowrap|dimension,{{citation

| last1 = Jaromczyk | first1 = J. W.

| last2 = Kowaluk | first2 = M.

| doi = 10.1016/0166-218X(91)90069-9

| issue = 2

| journal = Discrete Applied Mathematics

| pages = 181–191

| title = Constructing the relative neighborhood graph in 3-dimensional Euclidean space

| volume = 31

| year = 1991| doi-access = free

}}.{{citation

| last1 = Agarwal | first1 = Pankaj K. | author1-link = Pankaj K. Agarwal

| last2 = Mataušek | first2 = Jiří | author2-link = Jiří Matoušek (mathematician)

| contribution = Relative neighborhood graphs in three dimensions

| pages = 58–65

| title = Proc. 3rd ACM–SIAM Symp. Discrete Algorithms

| contribution-url = http://portal.acm.org/citation.cfm?id=139404.139416

| year = 1992}}.}} and for non-Euclidean metrics.{{citation

| last = O'Rourke | first = J. | author-link = Joseph O'Rourke (professor)

| doi = 10.1016/0031-3203(82)90070-X

| issue = 3

| journal = Pattern Recognition

| pages = 189–192

| title = Computing the relative neighborhood graph in the L_1 and L_\infty metrics

| volume = 15

| year = 1982}}.{{citation

| last = Lee | first = D. T. | author-link = Der-Tsai Lee

| doi = 10.1016/0031-3203(85)90023-8

| issue = 5

| journal = Pattern Recognition

| pages = 327–332

| title = Relative neighborhood graphs in the L_1-metric

| volume = 18

| year = 1985}}. Computing the relative neighborhood graph, for higher-dimensional point sets, can be done in time O(n^2).

Related graphs

The relative neighborhood graph is an example of a lens-based beta skeleton. It is a subgraph of the Delaunay triangulation. In turn, the Euclidean minimum spanning tree is a subgraph of it, from which it follows that it is a connected graph.

The Urquhart graph, the graph formed by removing the longest edge from every triangle in the Delaunay triangulation, was originally proposed as a fast method to compute the relative neighborhood graph.{{citation

| last = Urquhart | first = R. B.

| doi = 10.1049/el:19800386

| issue = 14

| journal = Electronics Letters

| pages = 556–557

| title = Algorithms for computation of relative neighborhood graph

| volume = 16

| year = 1980}}. Although the Urquhart graph sometimes differs from the relative neighborhood graph{{citation

| last = Toussaint | first = G. T. | author-link = Godfried Toussaint

| doi = 10.1049/el:19800611

| issue = 22

| journal = Electronics Letters

| page = 860

| title = Comment: Algorithms for computing relative neighborhood graph

| volume = 16

| year = 1980}}. Reply by Urquhart, pp. 860–861. it can be used as an approximation to the relative neighborhood graph.{{citation

| last1 = Andrade | first1 = Diogo Vieira

| last2 = de Figueiredo | first2 = Luiz Henrique

| contribution = Good approximations for the relative neighbourhood graph

| title = Proc. 13th Canadian Conference on Computational Geometry

| contribution-url = https://web.archive.org/web/20190328183825/https://rutcor.rutgers.edu/~dandrade/papers/ug.pdf

| year = 2001}}.

References