Computes the blur of an image by convolving a source image with a blurring
kernel. The blurring kernel is approximately a circle. A point is included
in the circle if its center is no more than Radius units from the center of
the blurring kernel. The weights of the elements in the circle are all
equal and sum to unity. If the size of the blur kernel (governed by the
Radius parameter) is zero then no blurring will take place. If the amount of
blurring (governed by the Blur parameter) is zero then no blurring will take
place. This blurring method is described in:
Digital Image Processing, Gonzales, R.C., Wintz, P.,
Addison Wesley, Second Edition, 1987, pp 163--173.
Port: Img In
Amount of blur.
Size of blur kernel.
Port: Img Out
Computation time is directly proportional to the size of the blur
kernel, which is a multiple of the Radius squared. If Radius is set to
be very large the module will run slowly. Also the amount of memory
required to blur the image is directly proportional to the size of the
kernel, thus if the kernel is very large then it is possible to run
out of available memory.
© The Numerical Algorithms Group Ltd, Oxford UK. 2000