基于麦克劳林展开与PCNN的医学图像融合  

Medical image fusion based on Maclaurin expansion and PCNN

作　　者：殷喆 高媛[1] 秦品乐[1] 刘朋伟 王丽芳[1] YIN Zhe;GAO Yuan;QIN Pinle;LIU Pengwei;WANG Lifang(College of Big Data,North University of China,Shanxi Provincial Key Laboratory of Biomedical Imaging and Imaging Big Data,Taiyuan 030051,Shanxi,China)

机构地区：[1]中北大学大数据学院,山西省生物医学成像与影像大数据重点实验室,山西太原030051

出　　处：《微电子学与计算机》2021年第12期47-53,共7页Microelectronics & Computer

基　　金：山西省自然科学基金(201901D111152)。

摘　　要：针对传统的基于多尺度变换方法使用的单一特征面忽略了曲线、边缘等互补特征的问题,提出了一种基于麦克劳林展开与高斯同态滤波增强相结合的脉冲耦合神经网络(PCNN)医学图像融合方法.首先通过麦克劳林展开将源图像分解为偏差分量和多级能量分量(以下分解到三级),再将三级能量分量进行高斯同态滤波增强获得增强的三级能量分量子图;然后使用自适应PCNN模型分别对偏差分量和多级能量分量进行融合,利用加权平均规则融合二级能量分量融合子图和增强的三级能量分量融合子图得到增强的能量分量融合图;最后反向麦克劳林展开获得融合图像.实验结果显示,该方法在图像清晰度、细节信息保留程度以及图像融合质量等方面与其他方法相比较更具有普适性.To address the problem that traditional multi-scale transform-based methods use single features while neglecting complementary features such as curves and edges.A Pulse Coupled Neural Network medical image fusion method based on a combination of Maclaurin expansion and Gaussian homomorphic filtering enhancement is proposed.The source image is firstly decomposed into deviance components and multi-level energy components(hereafter decomposed to three levels)by Maclaurin expansion,and then the three-level energy components are enhanced by Gaussian homomorphic filtering to obtain the enhanced three-level energy component sub-map The experimental results show that the method is more universal than other methods in terms of image sharpness,detail information retention and image fusion quality.

关 键 词：麦克劳林展开 高斯同态滤波增强 脉冲耦合神经网络 医学图像融合 加权平均规则 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]