pGPx (version 0.1.4)

dtt_fast: Rcpp implementation of Felzenszwalb distance transfom

Description

Rcpp wrapper for the distance transform algorithm described in Felzenszwalb and Huttenlocher (2012)

Usage

dtt_fast(x)

Value

A matrix of size \(n x m\) containing the distance transform result. Note that this function does not perform any checks on x.

Arguments

x

matrix of booleans of size \(n x m\) representing a (binary) image

Author

Pedro Felzenszwalb for the header files dt.h and misc.h that do the work, Dario Azzimonti and Julien Bect for the wrapper.

References

Felzenszwalb, P. F. and Huttenlocher, D. P. (2012). Distance Transforms of Sampled Functions. Theory of Computing, 8(19):415-428.

Examples

# Create an image with a square
nc = 256
nr = 256
xx = matrix(FALSE,ncol=nc,nrow=nr)
xx[(nr/16):(nr/16*15-1),nc/16]<-rep(TRUE,nr/16*14)
xx[(nr/16):(nr/16*15-1),nc/16*15]<-rep(TRUE,nr/16*14)
xx[nr/16,(nc/16):(nc/16*15-1)]<-rep(TRUE,nc/16*14)
xx[nr/16*15,(nc/16):(nc/16*15-1)]<-rep(TRUE,nc/16*14)
# Compute Distance transform
zz<- dtt_fast(xx)
# \donttest{
# Plot the results
image(xx,col=grey.colors(20), main="Original image")
image(zz,col=grey.colors(20), main="Distance transform")
# }