摘要
针对多传感器图像融合问题,提出了一种基于非下采样轮廓波变换域感受野模型的图像融合方法.首先,采用非下采样轮廓波变换对源图像进行多尺度、多方向稀疏分解;然后,对低频子图像采用改进型感受野模型进行融合,高频子图像则采用自适应Unit-Fast-Linking脉冲耦合神经网络模型进行融合;最后,将各子图像进行非下采样轮廓波逆变换,得到最终融合图像.仿真实验表明了所提出方法的有效性.
To the multi-sensor image fusion problem,a technique for image fusion based on non-subsampled contourlet transform(NSCT) domain receptive field model is presented.Firstly,by using NSCT,multi-scale and multi-direction sparse decomposition of source images are performed.Then,an improved receptive field model is utilized to achieve the fusion of the low frequency sub-images.In addition,the course of the high frequency sub-images fusion can be completed by using the model of adaptive unit-fast-linking pulse coupled neural network.Finally,the final fused image can be gained by adopting inverse NSCT to all sub-images.The simulation experimental results show the effectiveness of the proposed technique.
出处
《控制与决策》
EI
CSCD
北大核心
2011年第10期1493-1498,1503,共7页
Control and Decision
基金
国家自然科学基金项目(60773209)
关键词
图像融合
非下采样轮廓波变换
感受野
脉冲耦合神经网络
image fusion
non-subsampled contourlet transform
receptive field
pulse coupled neural network