Image fusion using self-constraint pulse-coupled neural network

Abstract

In this paper, an image fusion method using self-constraint pulse coupled neural network (PCNN) is proposed. A self-constraint restrictive function is introduced to PCNN neuron, so that the relation among neuron linking strength, pixel clarity and historical linking strength is adjusted adaptively. Then the pixels of original images corresponding to the fired and unfired neurons of PCNN are considered as target and background respectively, after which new fire mapping images are obtained for original images. Finally, the clear objects of original images are decided by the weighted fusion rule with the fire mapping images and merged into a new image. Experiment result indicates that the proposed method has better fusion performance than several traditional approaches. © 2010 Springer-Verlag.