FPGA Implementation of Pulse Coupled Neural Network on for Time Series of an Image

Abstract

Pulse Coupled Neural Network (PCNN) is biologically inspired neural networks, which has a good application in image processing, such as segmentation, enhancement, recognition, edge detection and so on. This paper presents a general VHDL modeling of PCNN, that is targeted for FPGA implementation, and can also be used with advantage for ASIC. First, the basic PCNN theory model is analyzed; and then the detail designed of each sub-module of the hardware is given; at last, the VHDL model is proved by comparing the time series output from FPGA simulation and that from theoretical calculation of the same image. The FPGA hardware implementation may be considered a platform for further, extended implementations and easily expanded into various applications.