摘要
提出了一种基于Contourlet变换的非下采样变换(Nonsubsampled ContourletTransform,NSCT)和脉冲耦合神经网络(Pulse Coupled Neural Network,PCNN)的可见光与红外图像融合算法。该算法首先对源图像进行NSCT分解,得到低频子带系数和各带通方向子带系数。然后对低频子带系数提出一种基于可见光与红外图像自身特性的加权平均融合方法,再对各带通子带系数提出基于PCNN的融合方法。最后经过NSCT逆变换得到融合图像。实验证明,该方法优于小波方法和传统的NSCT方法。
A visible and infrared image fusion algorithm based on Nonsubsampled Contourlet Trans-form (NSCT) and Pulse Coupled Neural Network (PCNN) is proposed. Firstly, the source images are decomposed by using NSCT and the low frequency subband coefficients and various bandpass directional subband coefficients are obtained. Secondly, a weighted averaging scheme based on the features of visible and infrared images is proposed for the low frequency subband coefficients. Then, the fusion method based on PCNN is proposed for each bandpass subband coefficient. Finally, the fused image is obtained after the inverse NSCT. The experimental result shows that the method is better than the wavelet-based fusion algorithms and the traditional NSCT-based fusion algorithms.
出处
《红外》
CAS
2013年第1期25-29,41,共6页
Infrared
关键词
图像融合
NSCT变换
脉冲耦合神经网络
可见光与红外图像
image fusion
nonsubsampled contourlet transform
pulse coupled neural network
visibleand infrared image