These functions provide standard mathematical morphology operations, which can be applied to array data with any number of dimensions. Binary and greyscale morphology is supported.
erode(x, kernel)dilate(x, kernel)
opening(x, kernel)
closing(x, kernel)
A morphed array with the same dimensions as the original array.
An object that can be coerced to an array, or for which a
morph method exists.
An array representing the kernel to be used. See
shapeKernel for functions to generate a suitable kernel.
Jon Clayden <code@clayden.org>
The erode function uses the kernel as an eraser, centring it on each
zero-valued pixel, which has the effect of eroding the extent of nonzero
areas. Dilation has the opposite effect, extending the nonzero regions in
the array. Opening is an erosion followed by a dilation, and closing is a
dilation followed by an erosion, using the same kernel in both cases.
If the kernel has only one unique nonzero value, it is described as
``flat''. For a flat kernel, the erosion is the minimum value of x
within the nonzero region of kernel. For a nonflat kernel, this
becomes the minimum value of x - kernel. Dilation is the opposite
operation, taking the maximum within the kernel.
morph for the function underlying all of these
operations, kernels for kernel-generating functions,
binarise for binarising an array, and
gaussianSmooth for smoothing. The EBImage
Bioconductor package also supplies functions to perform these operations,
and may be slightly faster, but only works in two dimensions.
x <- c(0,0,1,0,0,0,1,1,1,0,0)
k <- c(1,1,1)
erode(x,k)
dilate(x,k)
Run the code above in your browser using DataLab