基于Curvelet的彩色癌细胞分割新方法

New multicolor cancer cells segmentation method based on Curvelet

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。

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.