A nonlinear wave metric and its CNN implementation for object classification

In this paper a nonlinear wave metric is introduced for object classificati on. It is shown that the choice of a metric is a nontrivial problem since i t is easy to give examples when well-known distance measures, such as Hammi ng, Hausdorff, and Nonlinear Hausdorff metrics are completely inadequate fo r this classification. As an alternative a generalized theorem is proposed that includes the previous metrics as special cases. It is based on nonline ar wave propagation and defines a computational framework that is well-suit ed for parallel array processors. In this study we investigate different Ce llular Neural Network (CNN) architectures and solutions for the proposed me tric and analyze its VLSI implementation complexity.