摘要
仅采用小波变换技术融合CT/MRI医学图像时,只单次剔除CT/MRI医学图像不重要信息,残留大量冗余信息。为此,结合小波变换和Zernike不变矩方法处理CT/MRI医学图像,基于Zernike不变矩边缘检测算法构建较为理想的阶跃边缘模型,融合修正的CT/MRI医学图像放大效应后,通过Zernike不变矩检测图像亚像素边缘,首次剔除部分不重要信息;在此基础上采用小波变换方法将CT/MRI医学图像分割成3N个高频子带和1个低频子带,再次剔除CT/MRI医学图像中的不重要信息,最后依据图像区域方差值确定融合值,实现多个源CT/MRI医学图像信息融合。经过实验分析发现,融合后的CT/MRI图像能精准体现融合前图像信息,清晰度显著高于融合前图像。
When only the wavelet transform technology is used to fuse CT/MRI medical images,and the unimportant information in CT/MRI medical images is removed only once,a large amount of redundant information remains in the images.For this reason,the wavelet transform is combined with Zernike moment invariant method to process CT/MRI medical images,and an ideal step edge model is constructed based on Zernike invariant moment edge detection algorithm.After fusing the magnification effect of the modified CT/MRI medical images,the sub-pixel edges of the images are detected with Zernike moment invariant method,and some unimportant information is removed for the first time.On this basis,the wavelet transform method is used to segment CT/MRI medical images into 3 N high frequency sub-bands and one low frequency sub-band,and then the unimportant information in CT/MRI medical images is eliminated again.Finally,the fusion value is determined according to the variance value of image region to realize multi-source CT/MRI medical image information fusion.The experimental results show that the fused CT/MRI images can accurately reflect the image information before fusion,and the clarity is significantly higher than that before fusion.
作者
李晶晶
LI Jingjing(Department of Information and Control Engineering,Shenyang Urban Construction University,Shenyang 110167,China)
出处
《现代电子技术》
北大核心
2019年第21期68-72,共5页
Modern Electronics Technique
基金
2019年度校级科研发展基金项目(XKJ2019007)~~
关键词
CT/MRI医学图像
图像处理
图像分割
边缘检测
信息剔除
信息融合
CT/MRI
Medical image
image processing
image segmentation
edge detection
information removal
information fusion
