摘要
近年来,各向异性扩散的方法在图像去噪中广泛运用.然而,对于乘性噪声和低对比度图像,传统的各向异性方法易造成边缘扩散.提出了一种新的基于边界检测算子的超声图像抑制斑点的方法,其中边界检测算子结合了相位一致性和相位反对称的方法PCA.首先,结合相位一致性和相位反对称特征提取模型生成PCA,然后将PCA边界检测算子代替传统方法中的边界检测算子,从而生成新的基于边缘检测的各向异性扩散方法PCA-AD;最后,分别进行了模拟和实际超声图像实验.实验结果表明,PCA-AD方法在均值保持、减小方差和边缘定位方面胜过传统各向异性扩散方法.
In recent years,the anisotropic diffusion method has been widely used in the image noise reduction.However,conventional anisotropic diffusion is easy to cause edge diffusion for images corrupted with multiplicative noise and low contrast.This paper presents a new speckle reduction method for ultrasound images based on a new edge detector,namely the phase congruency and asymmetry detector(PCA).Firstly,PCA is developed from the phase congruency and symmetry model of feature detection.And then the edge detector in the traditional anisotropic diffusion is replaced by the PCA edge detector and the PCA edge detector-based anisotropic diffusion(PCAAD)is derived.Experiments are performed on both synthetic and real ultrasound images.The results show that the PCAAD meth od excels over traditional anisotropic diffusion method in terms of mean preservation,variance reduction and edge localization.
作者
季春红
余锦华
汪源源
Ji Chunhong;Yu Jinhua;Wang Yuanyuan(Department of Electronic Engineering,Fudan Unitersity,Shanghai 200433,China;Key laboratory of Medical Imaging Compuing and Computer Assisted Intervention of Shanghai,Shanghai 200433,China)
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2015年第S01期184-188,共5页
Chinese Journal of Scientific Instrument
基金
中国国家自然基金(81101049,61271071)
上海市浦江人才计划(12PJ1401200)
教育部博士点基金(20110071120019)项目资助
关键词
各向异性扩散
相位一致性
相位反对称
边缘检测算子
anisotropic diffusion(AD)
phase congruency
phase asymmetry
edge detection operator
