[WIMAGE, SUPPORT] = iat_inverse_warping(IMAGE, WARP, TRANSFORM, NX, NY, STR)

iat_inverse_warping implements the inverse warping of IMAGE. The coordinates defined by the combinations of NX,NY row-vectors are projected through WARP thus resulting in new subpixel coordinates. The intensities of new coordinates are computed via interpolation of IMAGE. For valid interpolation methods see the Matlab function interp2.

Input arguments:

IMAGE the input image that must be warped
WARP the warp transform
TRANSFORM the type of transformation. Valid string: {‘translation’, ‘euclidean’,’affine’,’homography’}
NX the row-vector with x-coordinates of the horizontal side of the region of interest  (i.e. [xmin:xmax])
NY the row-vector with y-coordinates of the vertical side of the region of interest, (i.e. [ymin:ymax])
STR (optional) the string that corresponds to interpolation method: ‘linear’, ‘cubic’ etc (for details see the Matlab function interp2), Default: ‘linear’


Output arguments:

WIMAGE the warped (interpolated) image
SUPPORT a binary double image that shows which areas in WIMAGE come from the support area of IMAGE and which from outside the borders


Back to Index