Applies Gaussian smoothing to a (multidimensional) image.
Smooths the n-dimensional array I with a n-dimensional gaussian with standard deviations
specified by sigmas. This operation in linearly seperable and is implemented as such.
INPUTS
I - imput image
sigmas - either n dimensional or 1 dimensional vector of standard devs
- if sigmas(n)<=.3 then does not smooth along that dimension
shape - [optional] shape to use in convolution [default == 'full']
radius - [optional] radius in units of standard deviation [default == 2.25]
OUTPUTS
L - smoothed image
filters - actual filters used, cell array of length n
DATESTAMP
29-Sep-2005 2:00pm