We consider the problem of segmenting a digitized image consisting of two u
nivariate populations. Assume a priori knowledge allows incomplete assignme
nt of voxels in the image, in the sense that a fraction of the voxels can b
e identified as belonging to population II0, a second fraction to II1, and
the remaining fraction have no a priori identification. Based upon estimate
s of the short length scale spatial covariance of the image, we develop a m
ethod utilizing indicator kriging to complete the image segmentation.