Wilson-Cowan neural network in image analysis

ABSTRACT The neural network model based on the theory proposed by Wi1son—owan has been simulated using digitizedreal images. The Wilson—Cowan net can operate in different modes depending on the parameter selection, and it isshown to store images in reduced form and to recognize edges of object. Examples how the net process the inputimages are shown. Due to large number of neurons in this model, the preferable technique to simulate it should beparallel processing one. Optics serve highly parallelism and we propose a basic hybrid—optical processing unit for the Wilson—Cowan net. 1. INTRODUCTION Image analysis research has been done since early 60's. The vast majority of the work on the field is based onthe heuristic, application oriented methods. However, there are different approaches to develope these methods.A basic approach is to think the image as a two-dimensional function and to solve the problems by mathematicaltools, e.g. edge detection by second derivative. This approach leads often into the linear image analysis. Another,nowadays actively studied approach is based on the (biologically motivated) neural networks. Here, the image is aninput to the network of simple processing units, neurons. The topology and behaviour of the net is often derivedfrom the knowledge about biological neural networks. These nets have e.g. properties like memory, recognition ofa specific pattern, and they can learn (modify their state) from examples.The Wilson -