A new approach for automated image segmentation based on simplified PCNN

Pulse-coupled neural network (PCNN) is a novel neural network,which has been widely used in image segmentation.However,there are still some limitations,such as trial-and-error parameter settings and manual selection of the optimal results.This paper puts forward a new method based on the simplified PCNN model for automatic image segmentation.By calculating the uniformity measure of the corresponding image at each process of iteration,the optimal segmentation result is obtained when the maximum value of the uniformity measure is achieved.Experimental results show that the proposed method can automatically achieve better segmentation result and has a common adaptability.