Modelling of a Chaotic System Motion in Video with Artificial Neural Networks

Abstract

In this study a chaotic motion is modelled by artificial neural networks which can be created again. Chaotic signals can occur in many fields like communication, encryption, nance, health, natural affairs. Artificial neural networks, fuzzy models can be used to provide a mathematical form and predict these types of signals as well. In this study, as an example of the motion which was modelled, there might be movement of a planet orbit, movements of balls on a billiard table, inverted pendulum or phase diagrams of such systems. However, for chaotic motion, a modified novel Lorenz system's phase diagram in literature was preferred. Object detection for motion which is sequential images of the video was obtained by image processing techniques so this process gives object coordination in the image. Artificial neural networks model which was called NAR structure was constructed and it has trained by these position information with the backpropagation algorithm. Subsequently, this NAR model which is artificial neural networks were tested and it was tried to get chaotic motion videos again. As a result, an object, which can be detected with image processing or other techniques, could be detected and traced. So, by using object information, which could be chaotic motion, could be modelled with artificial neural networks, instead of mathematically equations.

Keywords

• chaos
• chaotic systems
• artifical neural networks
• image processing