One method of region labeling involves examining each pixel in a mapping and comparing its value to that of its neighbors. If its value is ``close enough'' (as determined by some distance function) then it is determined as being in the same region as that neighbor. An equivalence table is maintained for a second pass that merges any regions determined to be the same (as in the case of labeling a Y-shaped region). As in the convolution operation, this method of region labeling is directly extendible to high-dimensional data by simply checking all N-dimensional neighbors. The pavement distress data shown in Figure 5a revealed 6 distinct regions as a result of this operation.