edgecluster {edci}R Documentation

Edge detection in noisy images

Description

edgecluster is a simple combination of edgepoints and oregMclust. It just passes the results of edgepoints to oregMclust.

Usage

  edgecluster(data, h1n, h2n,
              maxval,
              bw         = max(h1n,h2n)/qnorm(0.975),
              asteps     = 4,
              estimator  = "M_median",
              kernel     = "gauss",
              score      = "gauss",
              sigma      = 1,
              kernelfunc = NULL)

Arguments

data See description of edgepoints.
h1n, h2n See description of edgepoints.
asteps See description of edgepoints.
estimator See description of edgepoints.
kernel See description of edgepoints.
score See description of edgepoints.
sigma See description of edgepoints.
kernelfunc See description of edgepoints.
maxval See description of eplist.
bw See description of oregMclust. See default gives the observations within the 'edgecluster-window' 95% of the weight.

Value

Return value is a list of two numerical matrices. The first matrix contains the results of oregMclust, which are the 'edgeclusters'. The second matrix contains the result of edgepoints.

Author(s)

Tim Garlipp, garlipp@mathematik.uni-oldenburg.de

See Also

edgepoints, oregMclust, eplist

Examples

  y <- matrix(rep(0,60*60),nrow=60)
  y[21:40,21:40]<-1
  y <- y + matrix(rnorm(60*60,0,0.2),nrow=60)
  image(y,col=gray(seq(0,1,1/255)))

  ec <- edgecluster(y,0.05,0.05,0.7,
                    estimator="M_median",kernel="gauss")
  plot(bestMclust(ec[[1]],4),ec[[2]],xlim=c(0,1),ylim=c(0,1))

[Package edci version 1.0-1 Index]