An important problem in pattern recognition is determining the shape o
f a pattern class from its sampled points. We have reported earlier a
procedure in this regard which provides multivalued shape of a pattern
class in R-2. In the present article, an extension of the procedure t
o higher dimensional space (N) has been suggested. The effectiveness o
f the extended version has been demonstrated on some artificially gene
rated pattern classes (in R-3). The convergence of the estimated shape
to the original set has also been verified using two different metric
s.