| 引用本文： |
-
黄文明,邓珍荣,计冬华.基于支持向量机的红细胞彩色图像分割算法[J].广西科学院学报,2008,24(4):287-290. [点击复制]
- HUANG Wen-ming,DENG Zhen-rong,JI Dong-hua.Segmentation of Blood Cells Image Based on Support Vector Machines[J].Journal of Guangxi Academy of Sciences,2008,24(4):287-290. [点击复制]
|
|
| 摘要: |
| 针对细胞图像分割中红细胞目标提取和重叠红细胞分割两个难点,提出一种基于支持向量机(SVM)的红细胞彩色图像分割算法,并通过实验对算法进行验证。该算法利用SVM对原始图像进行红细胞提取,把原始细胞分割成红细胞和背景两类目标区域,然后使用改进距离标记的分水岭算法对红细胞区域进行重叠分割。算法选择线性不可分的SVM模型和核函数RBF(C=1,ξi=0.2)时能够较好的分割红细胞彩色图像。 |
| 关键词: 红细胞图像 支持向量机 分割 |
| DOI： |
| 投稿时间：2008-10-12 |
| 基金项目:2006年国家级火炬计划项目(2006GH041397)资助。 |
|
| Segmentation of Blood Cells Image Based on Support Vector Machines |
|
HUANG Wen-ming, DENG Zhen-rong, JI Dong-hua
|
| (Computer and Control College, Guilin University of Electronic Technology, Guilin, Guangxi, 541004, China) |
| Abstract: |
| Aiming at the difficulties of cells image segmentation which involve the blood image target extraction and segmentation of clustering blood cells, a segmentation algorithm is presented for blood cells image based on the Support Vector Machines (SVM). The algorithm begins with using SVM to extract the blood cells from the original image, along with dividing the image into two target areas of blood cells and background, followed by applying the improved distance marked watershed method to separate the blood cells from the overlapping areas. The choice of non-linear dividable SVM types and kernel RBF function at (C=1,ξi=0.2) results in satisfactory blood cells image segmentation. |
| Key words: blood cell image Support Vector Machines separation |
