Color Image Segmentation Based On Dynamic Model and Power Spectrum

摘要

In this paper, the locally coupled neural oscillator model and power spectrum are introduced to solve the color image segmentation problem. Firstly the single neuron oscillator model is developed, the model parameter is set so as to create limit circle simulating the activity of neuron. Then the locally coupled network model is developed, taking R, G, Bas external input separately, the network output three dynamic orbits representing R, G, B for each image point. Lastly, the power spectrum of the orbits is calculated for each image point, the color image is segmented using the calculated power spectrum. The performance is comparable to the results from other segmentation methods.