剪切波与脉冲耦合网络结合的医学图像融合  

Medical Image Fusion Combining Shear Wave and Pulse Coupling Network

作　　者：刘一男 张荣国[1] 李建伟[1] 王晓[1] 胡静[1] LIU Yi-nan;ZHANG Rong-guo;LI Jian-wei;WANG Xiao;HU Jing(School of Computer Science and Technology,Taiyuan University of Science and Technology,Taiyuan 030024,China)

机构地区：[1]太原科技大学计算机科学与技术学院,太原030024

出　　处：《太原科技大学学报》2021年第5期361-366,共6页Journal of Taiyuan University of Science and Technology

基　　金：国家自然基金(51375132);山西省自然科学基金(201801D121134);太原科技大学博士科研启动基金(20202057)。

摘　　要：针对图像融合产生的边缘模糊、对比度偏低、重要细节信息保留不充分的问题,提出了一种新的非下采样剪切波变换域的医学图像融合方法。首先对源图像进行非下采样剪切波变换,获得其细节图和近似图;对于细节图,采用细节特征信息作为外部激励条件,刺激脉冲耦合神经网络以实现图像融合;对于近似图,采用基于视觉显著映射实现融合;最后,进行逆剪切波变换得到融合图像。实验表明,此方法能有效提高融合图像的对比度以及细节丰富度等重要信息,与5种代表性的融合方法相比,本方法具有较好的视觉效果,在图像客观评价指标方面具有一定优势。Aiming at the problems of image fusion such as blurred edges,low contrast and insufficient retention of important details,a new multi-mode medical image fusion method in non-subsampled shear-wave transform domain is proposed.Firstly,a non-subsampled shear wave transform is applied to the source image to obtain its detail image and approximate image.For the detail image,the detail characteristic information is used as the external excitation condition and the pulse coupled neural network is stimulated to achieve image fusion.For approximate graphs,visual salient mapping is used to achieve fusion.Finally,the fusion image is obtained by inverse shear wave transform.Experimental results show that this method can effectively improve the image contrast,detail richness and other important information,compared with five representative medical image fusion methods,this method has better visual effect,and has certain advantages in the objective evaluation index of image.

关 键 词：图像融合 医学图像融合 剪切波变换 脉冲耦合神经网络 视觉显著映射 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]