Chaotic attractor of the image

The chaotic attractor is discovered as a new feature for image recognition in this paper. The chaotic characteristic of a dynamic system is composed of a wavelet function and an image function. Analyze the chaotic characteristic by calculating the Lyapunov exponent and drawing the bifurcation diagram. Then it is found that the number of attractor points can be changed by adjusting the parameter of wavelet function, but the shape of the attractor barely changed when the parameter changed slightly. The experimental result shows that using the chaotic attractor to distinguish different human faces is measurable.