

浏览全部资源
扫码关注微信
南京工业大学 电子与信息工程学院,江苏 南京,211816
Received:23 March 2014,
Revised:11 April 2014,
Published Online:18 May 2014,
Published:05 February 2015
移动端阅览
李晓龙, 帅仁俊. 一种基于形态学的脑肿瘤分割[J]. 液晶与显示, 2015,30(1): 157-162
LI Xiao-long, SHUAI Ren-jun. One kind of segmentation of brain tumors based on morphology[J]. , 2015,30(1): 157-162
李晓龙, 帅仁俊. 一种基于形态学的脑肿瘤分割[J]. 液晶与显示, 2015,30(1): 157-162 DOI: 10.3788/YJYXS20153001.0157.
LI Xiao-long, SHUAI Ren-jun. One kind of segmentation of brain tumors based on morphology[J]. , 2015,30(1): 157-162 DOI: 10.3788/YJYXS20153001.0157.
由于脑部结构复杂
存在精密纹理并且由于成像技术的限制
传统分水岭算法分割脑肿瘤很容易产生过分割问题.本文提出一种改进的分水岭算法解决此过分割现象.在应用分水岭算法之前对图像进行一些预处理:首先
应用形态学开闭重建重构原图像
滤除噪声和复杂细小的纹理;其次
对重建后的图像做多尺度形态学梯度
并对修改了标识符的梯度图像进行分水岭算法分割.实验结果表明:结构元素半径大小取17~22
肿瘤能够完全地分割出来.与传统的分水岭分割算法相比
本文所采用的算法很好地抑制了过分割问题
可以很容易地分割出肿瘤.
As the brain is complicated
which has precision texture
and imaging technologies is limited
there will be an over segmentation issue if the traditional watershed algorithm is used to segment brain tumor. As a matter of method
some pre-processing is proposed before applying the watershed segmentation: firstly
morphological opening and closing reconstruction is used to reconstruct the original image and to filter the noise and fine texture; secondly
the multi-scale morphological gradient is used in the reconstructed image
and the watershed segmentation algorithm is used in the gradient image of modified marker. Experiments results indicate that the tumor can be separated completely when structural element radius size is used from 17 to 22. Compared with the traditional watershed segmentation algorithm
this method alleviates the over segmentation problem
and tumor can be separated easily from the original image.
税午阳,周明全,耿国华. 磁共振颅脑图像的脑组织自动获取方法[J]. 软件学报, 2009,20(5):103-106. Shui W Y,Zhou M Q,Geng G H. Method of automatic acquisition of magnetic resonance image of The brain [J].Journal of Software,2009,20(5):103-106.(in Chinese)[2] Vincent L,Soille P.Watershed in digital spaces;An efficient algorithrm based on immersion simulations [J].IEEE Trans.on Pattern Analysis and Machine Intelligence,1991,13(6):583-589.[3] 张鹤,吴谨. 基于运动信息和标记多尺度分水岭的运动目标检测算法[J]. 液晶与显示, 2012,27(2):250-256. Zhang H,Wu J.Moving target detection algorithm based on motion information and multi-measure watershed [J].Chinese Journal of Liquid Crystals and Displays,2012,27(2):250-256.(in Chinese)[4] 赵文达,赵建,韩希珍,等. 基于变分偏微分方程的红外图像增强算法研究[J]. 液晶与显示, 2014,29(2):281-285. Zhao W D,Zhao J,Han X Z, et al.Infrared image enhancement based on varitional partial differential equations [J].Chinese Journal of Liquid Crystals and Displays,2014,29(2):281-285.(in Chinese)[5] 张强,王正林.精通MATLAB图像处理 [M].2版.北京:电子工业出版社,2012. Zhang Q,Wang Z L.Proficient in MATLAB Image Processing [M].2nd ed. Beijing: Electronics Industry Press,2012.(in Chinese)[6] 葛世国,王茂芝,刘娟娟,等. 基于标记的改进分水岭分割算法[J]. 四川理工学院学报:自然科学版 [J].2013,26(2):57-61. Ge S G,Wang M Z,Liu J J, et al.Improved watershed segmentation algorithm based on marker [J].Journal of Sichuan University of Science and Engineering:Natural Science [J].2013,26(2): 57-61.(in Chinese)[7] 冈萨雷斯.数字图像处理(matlab版) [M].北京:电子工业出版社,2013. Gonzalez.Digital Image Processing(MATLAB) [M].Beijing: Electronics Industry Press,2013.(in Chinese)[8] 韩敬华,何长涛, 张秋慧,等. 应用标记控制分水岭分割算法研究激光诱导玻璃体损伤形貌[J]. 光学精密工程, 2010,18(6):1387-1395. Han J H,He C S,Zhang Q J, et al.Morphological analysis of laser induced glass bulk damage by marker controlling watershed algorithm [J].Optics and Precision Engineering,2010,18(6): 1387-1395.(in Chinese)[9] 侯阿临,张利红,徐坤,等. 一种适合医学图像分割的改进分水岭算法[J]. 电视技术, 2007,31(11): 87-89. Hou A L,Zhang L H,Xu K,et al.An improved watershed segmentation algorithm for medical image [J].Video Engineering,2007,31(11):87-89.(in Chinese)[10] 程淑红,张典范,胡春海,等. 基于标记的多尺度分水岭视频目标分割算法[J]. 光电工程, 2010,37(4): 130-134. Cheng S H,Zhang D F,Hu C H,et al.Video object segmentation algorithm based on marker multi-measure watershed [J].Opto-Electronic Engineering,2010,37(4):130-134.(in Chinese)
0
Views
26
下载量
2
CSCD
Publicity Resources
Related Articles
Related Author
Related Institution
京公网安备11010802024621