An Adaptive Image Segmentation Method Based on a Modified Pulse Coupled Neural Network

Abstract

Pulse-coupled neural network (PCNN) based on Eckhorn's model of the cat visual cortex has great significant advantage in image segmentation. However, the segmented performance depends on the suitable PCNN parameters, which are tuned by trial so far. Focusing on the famous difficult problem of PCNN, this paper establishes a modified PCNN, and proposes adaptive PCNN parameters determination algorithm based on water region area. Experimental results on image segmentation demonstrate its validity and robustness.