Abstract
NonSubsampled Contourlet Transform (NSCT) has the characteristics of multi-scale, multi-directional, multi-resolution and shift invariance. Because of down sampling, traditional contourlet transform will cause Gibbs phenomenon, NSCT can overcome the disadvantage, obtaining better fusion image. Due to Pulse Coupled Neural Networks (PCNN) excellent biological characteristics, it has already been widely applied to image processing. In this paper, we will combining NSCT and PCNN, making full use of their advantages and applying to image fusion. The original image is by NSCT transform, getting the decomposition coefficients and then calculating its spatial frequency, input them to PCNN. According to the firing times select fusion coefficients. Experiment result has shown that, compared with the typical wavelet transform and contourlet transform method, the method proposed in this paper whether in subjective or objective is better than other methods. 1548-7741/Copyright © 2015 Binary Information Press
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
