Design and realization of cellular neural networks for optimization and image processing

Laith khalid Fayadh Al-Karbooly
Dr. Nasser N. Al-Ani

Abstract : Cellular Neural Network(CNN) used and tested in this work for solving many scientific problems in image processing, pattern recognition, and optimization by use of stored program with designed templates. As it is known, CNNs are locally interconnected arrays; these arrays consist of identical dynamical cells (neurons) that operate in parallel. The connection between cells is controlled by the designed template. These templates determine exclusively the CNN dynamic behavior. An Operational Transconductance Amplifier (OTA) circuit is proposed and used for modeling neurons in VLSI implementation of selected image processing and ptimization tasks. The OTA circuit generates an output current which is a function of the linear sum of a number of weighted inputs. The weight of each input is controlled by a bias voltage, in addition to the aspect ratio of the transistor of each cells. In the present work; the cell structure, dynamic range, and the stability analysis of CNN and TVCNN are discussed. Cellular neural network template is designed and tested to perform a number of image processing operation and optimization task using MATLAB 6.5. The templates are implemented as a set of unit programmable OTA. The ORCAD PSPICE 9.1 program is used to simulate circuits that are designed for optimization and image processing.