This chapter discusses the use of asynchronous cellular automata with controlled movement of active cells for image processing and recognition. A time-pulsed image description method is described. Various models and structures of cellular automata for transmitting active signals are presented. The image of the figure is binarized and an active signal moves along its edges. At every moment in time, the active cell of an asynchronous cellular automaton generates a pulse signal. The shape of the generated pulse sequence describes the geometric shape of a flat figure. Methods for describing images of individual plane figures, as well as a method for describing images consisting of many separate geometric objects, are proposed. Cellular automaton is considered as an analogue of the retina of the human visual canal. The circuitry structures of cells of such asynchronous cellular automata are presented, and the software implementation of the proposed methods is also performed. Methods allow one to classify individual geometric image objects.
TopRetina Analog Of Optic Channel Based On Cellular Automata
Modern studies of the physiological structure of the retina of the human visual canal have made it possible to determine the main structural elements and the relationships between them (Schachat, et al 2017; Fineman, 2018; Dowling, 2012). Structurally, the retina can be represented as a horizontal slice (Figure 1).
Figure 1. Simplified horizontal retinal cut model
The first layer of the retina consists of photoreceptors that respond to electromagnetic radiation in the visible part of the spectrum and convert it into electrical (chemical) signals coming to the following functional layers of the retina. Photoreceptors carry out primary processing of optical signals. The subsequent layers of the retina as the main discrete elements contain (Figure 1):
H – horizontal cells;
B – bipolar cells;
M – Mueller cells;
А – amacrine cells;
G – ganglion cells.
All these cells form layers and are interconnected in a certain way. In fact, the retina is a discrete multilayer structure that consists of several groups of homogeneous cells. These cells are intertwined with nerve fibers with many synaptic contacts. As a result of the passage of signals through layers of interwoven cells, and already with semantic content, they enter the ganglion layer. Ganglion cells carry out the collective function of signals from cells of previous layers. At the outputs of the ganglion cells, sequences of impulses are formed (Figure 2), which enter the optic nerve through the optic nerve to the cerebral cortex (Schachat, et al 2017; Fineman, 2018; Dowling, 2012).
Figure 2. Forms of electrical signals at the outputs of ganglion cells
At present, the authors are not aware of the results of studies that are devoted to a detailed study of the structure and forms of pulse sequences. The exact relationship between the optical signals at the photoreceptors and the pulse sequences at the outputs of G is not described.