Human perception is a complex nonlinear dynamics. Motivated by biological experimental findings, two networks of coupled chaotic elements for image segmentation are introduced in this chapter. In both models, time evolutions of chaotic elements that correspond to the same object in a given image are synchronized with one another, while this synchronized evolution is desynchronized with respect to time evolution of chaotic elements corresponding to other objects in the image. The first model is a continuous flow and the segmentation process incorporates geometrical information of input images; while the second model is a network of discrete maps for pixel clustering, accompanying an adaptive moving mechanism to eliminate pixel ambiguity. Computer simulations on real images are given.