基于自适应字典学习降噪改进的脑MRI图像分水岭精确分割算法研究
|
摘要
图像分割技术是图像处理领域非常活跃的研究课题.但目前还不完善尤其MRI高噪声图像分割还没有给出较好的分割算法,如:脑MRI图像的分割等等.分水岭算法和图像降噪算法在图像分割中有广泛的应用.本文将这两种方法结合起来,并用于脑MRI图像分割,取得较好的分割结果.本文详细论述了字典学习降噪算法的原理,提出一种字典学习降噪和分水岭算法相结合的脑MRI医学图像分割算法.采用字典学习降低原始图像噪声,然后利用形态学算法对降噪后的图像进行形态学处理,通过形态学知识提取图像边界.利用图像的几何特征,去除非目标区域,再采用分水岭变换进行图像分割,并通过脑MRI图像验证了此方法的优势.实验结果进一步验证了其可行性.
Image segmentation technique is research topic of very active in the field of image processing. However,not yet perfect,especially MRI noisy image segmentation is not gives better algorithm of segmentation,such as:MRI Image Segmentation of brain,etc. Watershed algorithm and image noise reduction algorithm is widely used in image segmentation. In this paper,the two methods are combined,and for Segmentation of brain MRI Image,and obtain better results of segmentation. In this paper discusses principles of the algorithm of dictionary learning noise reduction.An image segmentation algorithm proposed of dictionary learning noise reduction and watershed algorithm combining.First,that original image reduces noise by dictionary learning; secondly,that use algorithms of morphological processing image of noise reduction after; finally,that extracted image boundary by morphological knowledge. We use geometric features of the image,and remove non- target area,and using watershed transformation segmentation of the image,and verify the advantages of the method by MRI image of brain. The results further validate the feasibility
Image segmentation technique is research topic of very active in the field of image processing. However,not yet perfect,especially MRI noisy image segmentation is not gives better algorithm of segmentation,such as:MRI Image Segmentation of brain,etc. Watershed algorithm and image noise reduction algorithm is widely used in image segmentation. In this paper,the two methods are combined,and for Segmentation of brain MRI Image,and obtain better results of segmentation. In this paper discusses principles of the algorithm of dictionary learning noise reduction.An image segmentation algorithm proposed of dictionary learning noise reduction and watershed algorithm combining.First,that original image reduces noise by dictionary learning; secondly,that use algorithms of morphological processing image of noise reduction after; finally,that extracted image boundary by morphological knowledge. We use geometric features of the image,and remove non- target area,and using watershed transformation segmentation of the image,and verify the advantages of the method by MRI image of brain. The results further validate the feasibility
引文
[1]孙黎明.图像稀疏去噪算法的并行改进研究[D].重庆大学,2011.
[2]姜鹏飞.基于稀疏表示与字典学习的图像去噪算法研究[D].西安电子科技大学,2011.
[3]尹忠科,解梅,王建英.基于稀疏分解的图像降噪[J].电子科技大学学报,2006,35(6):20-21.
[4]张春梅,尹忠科,肖明霞.基于冗余字典的信号超完备表示与稀疏分解[J].科学通报,2006,51(6):628-633.
[5]汪胜前.图像的小波稀疏表示及收缩降噪算法[D].上海交通大学博士论文,2002.
[6]GRAU V,MEWES AUJ,ALCANIZ M.Improved watershed transform for medical image segmentation using prior information[J].IEEE Transaction on Medical Imaging,2004,23(4):447-458.
[7]Bieniek A,Moga A.An efficient watershed algorithm based on connects components[J].Pattern Recognition,2000,(3):907-716.
[8]HIEU T,MARCEL M,REIN V.Water snakes:Energy driven watershed segmentation[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,2003,25(3):330-342.
[9]王金凤.基于数学形态学的彩色图像边缘检测[D].燕山大学,2010.
[10]潘婷婷.数学形态学和分水岭算法在遥感图像目标识别中的应用研究[D].江南大学,2008.
[11]许晶.VDRL图像处理研究及监测报告管理系统的设计与实现[D].哈尔滨理工大学,2008.
[12]陈婷婷.采用模糊形态学和分水岭算法的图像分割研究[D].西南大学硕士论文,2008.
[13]Mallat S and Zhang Z.Matching pursuit with time-frequency dictionaries.IEEE Transactions on Signal Processing[J],1993,41(12):3397-3415.
[2]姜鹏飞.基于稀疏表示与字典学习的图像去噪算法研究[D].西安电子科技大学,2011.
[3]尹忠科,解梅,王建英.基于稀疏分解的图像降噪[J].电子科技大学学报,2006,35(6):20-21.
[4]张春梅,尹忠科,肖明霞.基于冗余字典的信号超完备表示与稀疏分解[J].科学通报,2006,51(6):628-633.
[5]汪胜前.图像的小波稀疏表示及收缩降噪算法[D].上海交通大学博士论文,2002.
[6]GRAU V,MEWES AUJ,ALCANIZ M.Improved watershed transform for medical image segmentation using prior information[J].IEEE Transaction on Medical Imaging,2004,23(4):447-458.
[7]Bieniek A,Moga A.An efficient watershed algorithm based on connects components[J].Pattern Recognition,2000,(3):907-716.
[8]HIEU T,MARCEL M,REIN V.Water snakes:Energy driven watershed segmentation[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,2003,25(3):330-342.
[9]王金凤.基于数学形态学的彩色图像边缘检测[D].燕山大学,2010.
[10]潘婷婷.数学形态学和分水岭算法在遥感图像目标识别中的应用研究[D].江南大学,2008.
[11]许晶.VDRL图像处理研究及监测报告管理系统的设计与实现[D].哈尔滨理工大学,2008.
[12]陈婷婷.采用模糊形态学和分水岭算法的图像分割研究[D].西南大学硕士论文,2008.
[13]Mallat S and Zhang Z.Matching pursuit with time-frequency dictionaries.IEEE Transactions on Signal Processing[J],1993,41(12):3397-3415.