Up: Correlation Analysis Tasks
Computes the cross histogram of two input images/cubes. The output has
the same rank as the input arrays : 2 if images as input, 3 if cubes.
Both inputs must have the same rank. If the input cubes are 3D data
cubes, the third axis must be the same for both. If n1 and n2 are the
numbers of bins for the first and the second input cube respectively,
the result has dimensions [n1,n2] or [n1,n2,nchan] where nchan is the
third dimension of the input cubes. The value at a given (I,J) is thus
the number of pixels in the input images that have the value correspond-
ing to slot I in the first image and to slot J in the second one. For
cubes, this is done in each plane. You can then sum (see SIC\COMPUTE) to
have the cross histrogram for the whole cube.
If the tolerance on the blanking value is positive, the task takes into
account the blanking values of the input cubes.
The output image can be used as input to task REGRESSION to evaluate
some statistical parameters of the correlation. See also HISTO_CLOUD for
a slightly different information.