| 基于Curvelet的彩色癌细胞分割新方法 |
| |
| 引用本文： | 李健,牛振山. 基于Curvelet的彩色癌细胞分割新方法[J]. 计算机工程与设计, 2012, 33(2): 654-657 |
| |
| 作者姓名： | 李健 牛振山 |
| |
| 作者单位： | 陕西科技大学电气与信息工程学院,陕西西安,710021 |
| |
| 基金项目： | 陕西省科技计划基金项目，陕西省教育厅专项科研计划基金项目，温州市科技合作基金项目 |
| |
| 摘 要： | 针对彩色图像分割问题,将Curvelet变换与SVM理论相结合,形成了有效的彩色图像分割新方法.通过曲波变换将彩色图像分解到各通道,用Mean Shift找到各通道上特征图像的模式点,再用模式点周围的样本训练SVM,用训练好的SVM对各通道样本进行精确分类,把所有通道滤波后的图像进行重构,使癌细胞凸显并二值化.该方法可以快速,精确地定位到多目标物边界.通过MATLAB进行仿真实验,表明了该方法的有效性.
|
| 关 键 词： | 机器视觉 曲波滤波 均值漂移 多分类支持向量机 彩色图像分割 |
New multicolor cancer cells segmentation method based on Curvelet |
| |
| LI Jian , NIU Zhen-shan. New multicolor cancer cells segmentation method based on Curvelet[J]. Computer Engineering and Design, 2012, 33(2): 654-657 |
| |
| Authors: | LI Jian NIU Zhen-shan |
| |
| Affiliation: | (College of Electricity and Information Engineering,Shaanxi University of Science and Technology,Xi’an 710021,China) |
| |
| Abstract: | For the problem of multicolor images segmentation,a new multicolor image segmentation approach is proposed based on SVM theory combined with Curvelet transform.First,the image is decomposed into multi-channel by Curvelet transform.Secondly,each channel’s feature image is filtered by Mean Shift combined with SVM theory to find the singular points.Finally,the filtered images of all channels are reconstructed to make defects prominent,and the binary image is obtained by threshold.The multi-objects boundary is located fast and accurately.The effectiveness of method is verified by MATLAB simulation experiments. |
| |
| Keywords: | machine vision Curvelet filtering M-SVM Mean Shift multicolor images segmentation |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
