MATLAB Examples

nanimresize documentation

The nanimresize function resizes an image using the Image Processing toolbox function imresize, but first fills NaNs to prevent missing data along NaN boundaries.



This function requires Matlab's Image Processing Toolbox. This function includes John D'Errico's inpaint_nans function as a subfunction. Thanks John.


B = nanimresize(A,scale)
B = nanimresize(A,[NumRows NumCols])
[Y,NewMap] = nanimresize(X,Map,scale)
[Y,NewMap] = nanimresize(X,Map,[NumRows NumCols])


B = nanimresize(A, scale) returns an image that is scale times the size of A, which is a grayscale, RGB, or binary image.

B = nanimresize(A, [NumRows NumCols]) resizes the image so that it has the specified number of rows and columns. Either |NumRows or NumCols may be NaN, in which case imresize computes the number of rows or columns automatically in order to preserve the image aspect ratio.

[Y, NewMap] = nanimresize(X, Map, scale) resizes an indexed image.

[Y, NewMap] = nanimresize(X, Map, [NumRows NumCols]) resizes an indexed image.

To control the interpolation method used by imresize, add a method argument to any of the syntaxes above, like this:

     nanimresize(A, scale, method)
     nanimresize(A, [NumRows NumCols], METHOD),
     nanimresize(X, MAP, M, METHOD)
     nanimresize(X, MAP, [NUMROWS NUMCOLS], METHOD)

method can be a string naming a general interpolation method:

     'nearest'    - nearest-neighbor interpolation
     'bilinear'   - bilinear interpolation
     'bicubic'    - cubic interpolation; the default method

method can also be a string naming an interpolation kernel:

     'box'        - interpolation with a box-shaped kernel
     'triangle'   - interpolation with a triangular kernel
                       (equivalent to 'bilinear')
     'cubic'      - interpolation with a cubic kernel
                       (equivalent to 'bicubic')
     'lanczos2'   - interpolation with a Lanczos-2 kernel
     'lanczos3'   - interpolation with a Lanczos-3 kernel

Finally, method can be a two-element cell array of the form {f,w}, where f is the function handle for a custom interpolation kernel, and w is the custom kernel's width. f(x) must be zero outside the interval -w/2 <= x < w/2. Your function handle f may be called with a scalar or a vector input.

You can achieve additional control over nanimresize by using parameter/value pairs following any of the syntaxes above. For example:

     B = nanimresize(A, SCALE, PARAM1, VALUE1, PARAM2, VALUE2, ...)

Parameters include:

     'Antialiasing'  - true or false; specifies whether to perform
                       antialiasing when shrinking an image. The
                       default value depends on the interpolation
                       method you choose.  For the 'nearest' method,
                       the default is false; for all other methods,
                       the default is true.
     'Colormap'      - (only relevant for indexed images) 'original'
                       or 'optimized'; if 'original', then the
                       output newmap is the same as the input map.
                       If it is 'optimized', then a new optimized
                       colormap is created. The default value is
     'Dither'        - (only for indexed images) true or false;
                       specifies whether to perform color
                       dithering. The default value is true.
     'Method'        - As described above
     'OutputSize'    - A two-element vector, [MROWS NCOLS],
                       specifying the output size.  One element may
                       be NaN, in which case the other value is
                       computed automatically to preserve the aspect
                       ratio of the image.
     'Scale'         - A scalar or two-element vector specifying the
                       resize scale factors.  If it is a scalar, the
                       same scale factor is applied to each
                       dimension.  If it is a vector, it contains
                       the scale factors for the row and column
                       dimensions, respectively.


Load some 0.25 degree sample data and resize it to 1 degree data:

load nanimresize_example.mat

shading flat
title 'original data'

Zr = imresize(Z,0.25);
lat_1deg = imresize(lat,0.25);
lon_1deg = imresize(lon,0.25);
shading flat
title 'imresize'

Zr_nan = nanimresize(Z,0.25);
shading flat
title 'nanimresize'

shading flat
title 'data lost by imresize'

Author Info

This function was written by Chad A. Greene of the University of Texas at Austin's Institute for Geophysics (UTIG), April 2016.

Although Chad wrote this function, it is heavily dependent on John D'Errico's inpaint_nans function, which is included as a subfunction here. Many thanks to John for writing and sharing such a fine function.