A new concept for invariant pattern recognition is presented that uses
object contour information. First, an angular signature of the object
contour is obtained by a nonlinear operation applied to two-dimension
al directional convolutions with a long, narrow kernel. The angular si
gnature function is normalized by either its area or its energy to ach
ieve quasi-invariance to scale. The resulting signature is then compar
ed with template signatures for the invariant recognition for which an
angular similarity measure is obtained from a one-dimensional correla
tion between the two signatures. Numerical experiments demonstrate tha
t the method discussed exhibits invariance to shift and angular orient
ation and quasi-invariance to scale.