Morphological image processing has been widely used to
process binary and grayscale images. To extend the concept to color
images, an ordering of the data is required. The solution is not unique,
because color spaces are not totally ordered and the ordering process is
not straightforward. In this work, two algorithms for color morphology are
proposed: A Mahalanobis-color-distance-based morphological ordering
algorithm, and a corrected componentwise morphological ordering algorithm.
Both algorithms implement the Mahalanobis color measure to replace
the angle-valued pixels by a scalar, and are based on a combination
of reduced and conditional ordering of the underlying data.