基于小波变换的图像融合研究  

Research on image fusion based on wavelet transform

作　　者：胡纪昌 董男 邵彤 尉红梅 范一峰 Hu Jichang;Dong Nan;Shao Tong;Wei Hongmei;Fan Yifeng(School of Medical Imaging,Hangzhou Medical College,Hangzhou,Zhejiang 310053,China)

机构地区：[1]杭州医学院医学影像学院,浙江杭州310053

出　　处：《计算机时代》2023年第2期109-110,114,共3页Computer Era

基　　金：国家级大学生创新创业项目“神经导航引导的脑肿瘤穿刺机器人研究”(202213023001)。

摘　　要：在导航手术过程中,图像引导起着关键作用,直接影响导航精度。CT图像能清晰显示颅骨和粘贴的标记点,却对脑组织的显示有限。MRI图像对脑内部软组织结构分辨高,但对头颅表面信息显示模糊。因此单一模态的图像对于导航手术引导有限。本文提出基于小波变化的图像融合方法,对CT和MRI图像进行了价值信息的有效融合,弥补了单幅图像信息不足的缺点。实验结果表明,该图像融合方法具有较强的鲁棒性,能够满足临床应用。Image guidance plays a key role in navigation surgery, which directly affects the navigation accuracy. CT images can clearly show the skull and labeled marks, but the display of brain tissue is limited. MRI images have a high resolution of the soft tissue structure in the brain, but show fuzzy information on the skull surface. Therefore, single mode images are limited in guiding navigation surgery. An image fusion method based on wavelet changes can effectively fuse the value information of CT and MRI images, and make up for the shortcoming of insufficient information of a single mode images. The experimental results show that the image fusion method has strong robustness and can meet the clinical application.

关 键 词：导航手术 CT MRI 小波变化 图像融合 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]