Implementation of binary mathematical morphology using discrete-time cellular neural networks

In this paper, the structures of the discrete-time cellular neural network (DTCNN) are modified for the purpose of implementation of binary mathematical morphology operations. The structure of a DTCNN for implementation of the four basic mathematical morphology operations: dilation, erosion, opening and closing are given. Two single layer DTCNNs are used to implement binary dilation and binary erosion, and two 2-layer DTCNNs are used to implement binary opening and binary closing, respectively. Simulation results are given.