The Study Of PCNN And Chaotic PCNN

With the studies and progress of the biological neurology, a kind of new artificial neural network model known as the third generation of artificial neural network—the pulse coupled neural network(PCNN) rises quietly. This network is different from the traditional artificial neural network. It is constructed by simulating the activities of the cat and dog's visual cortex neurons. And it is the simplification and approximation of the real neurons, so its prospect of application is very extensive. The properties inherited in PCNN, such as linking field and the dynamic threshold, make the approximate neurons fire simultaneously. And it is very close to the natures of visual cortex of small mammals. So it has gained widely applications in image processing.On the basis of introducing the basic principle of the pulse coupled neural network, this thesis summarizes the application of PCNN used in image processing, and simulate the arithmetic of image smoothing. It also introduces the chaos theory and chaotic neural network models. It tries to introduce the chaotic phenomenon in the the pulse coupled neural network .The main work of the thesis is as follows:1. Make a more comprehensive summary and comparison the different algorithm of the pulse coupled neural network appeared in recent years in image smoothing, image enhancement, image segmentation and so on several aspects, and analyzed its fit and unfit quality, it confirmed that PCNN used in image processing havs incomparable biology superiority than other networks.2. Describes the algorithm of PCNN used to reduce the noise of two-level images which are polluted by salt-and-pepper noise. The results of computer simulations show that visual effects of the restoration images by using PCNN are much better than those by using median filter. It also gives the research of the algorithm of image edge detection used PCNN. The thesis also analyzes the parameters in PCNN algorithm in detail, so that the understanding of PCNN is expanded further.3. Through changing resetting threshold value from constant into the sinusoidal oscillating, we get the chaotic pulse coupled neural network model with chaos phenomenon. We analyze the dynamic behavior of the single neuron and the condition of two coupled neurons keeping firing at the same time; the phase of firing time is simulated by classifying its behavior. The existence of chaos is confirmed by Lyapunov Exponents.