基于第二代Curvelet变换和特征量积的图像融合  

Image Fusion Based on the Second Generation Curvelet Transform and Feature Product

作　　者：胡学龙[1] 潘大伟[1] 陆慧敏[1,2] 张力峰 Seiichi SERIKAWA 

机构地区：[1]扬州大学信息工程学院,江苏扬州225127 [2]九州工业大学工学部

出　　处：《微电子学与计算机》2012年第11期126-129,共4页Microelectronics & Computer

基　　金：江苏省第七批六大人才高峰高层次人才项目(110)

摘　　要：文中提出了一种基于二代Curvelet和特征量积的图像融合算法.首先对融合图像进行快速离散Curvelet变换,在相应的尺度上由融合规则将Curvelet系数融合.采用平均策略融合低频系数;对于高频分量采用了小波的局部能量、局部梯度、局部的标准偏差组成的特征量积自适应地融合高频系数.最后利用Curvelet重构得到融合图像.对多聚焦图像和多光谱图像进行了实验.采用了标准差、平均梯度、信息熵作为评价标准,并与小波的融合结果做了比较..实验证明二代Curvelet所获得的融合结果更佳清晰、效果更佳且应用广泛.This paper puts forward a new image fusion method based on the second generation Curvelet and feature product.Firstly,the fast discrete Curvelet transform is performed on the original images to obtain the coefficients at the corresponding scale.Then the corresponding subband images by using different rules are fused to get the Curvelet coefficient.At last,the resulting image is obtained by inverse Curvelet transform.The new method was tested by multi-foucs images and multi-spectral images.The standard deviation,average gradient and comentropy are used to evaluate the experimental results.And comparisons with the results based on wavelet transform are also carried out.The test results show that the resulting images based on the second generation Curvelet transform are more clear and better.

关 键 词：图像融合 第二代CURVELET变换 特征量积 质量评价 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]