摘要
本文提出了一种新的基于小波变换的多传感器图像融合算法,在线性融合模型中引入了传感器噪声项,算法首先给出了基于期望融合图像局部方差最大的多传感器融合系数估计方法,然后利用多传感器图像和融合系数估计期望融合图像.最后的实验结果表明了本文图像融合方法具有明显的噪声抑制能力,能够得到更好的融合效果.
A new method based on wavelet transform for multi-sensor image fusion is presented in this paper. The sensor noise is introduced into linear fusion model. At first, the fusion coefficients of multiple sensors are given based on the maximization of local variance of the desired fusion image. Then, the desired fusion image is estimated by using the multi-sensor images and the fusion coefficients. The experiment results demonstrate that this method can greatly reduce sensor noise and achieve better fusion performance.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2005年第6期1008-1010,共3页
Acta Electronica Sinica
关键词
图像融合
小波变换
线性融合模型
Algorithms
Estimation
Image processing
Mathematical models
Optimization
Spurious signal noise
Wavelet transforms
