Connectivity analysis of multi-dimensional multi-valued images

Connectivity analysis (maximally connected component labeling plus optional geometric feature collection) has previously been applied to only 2-dimensional (2-D) binary valued images. By carefully examining the property of 6-connectivity, it is found that it may also be applied to 2-D multi-valued (m-ary) images, which, together with thresholders, recognizers or classifiers of multiple valued output, promises more efficient low level processing. The idea is further generalized to multi-dimensional spaces so that the connectivity analysis may be performed on n-D (with n ≥ 1) m-ary (with m ≥ 2) images. Formal definition of 6-connectivity in n-D space and a labeling algorithm is presented followed by a brief discussion of its potential applications.