摘要
基于全局信息的CV(Chan-Vese)模型不能有效分割灰度不均匀的图像,而图像局部信息更能反映目标图像的特征。在CV模型的基础上融入目标的局部信息,并在目标函数中使用相对熵度量最小化轮廓内外区域的相似度,以提高图像分割准确度和抑制图像噪声。实验结果表明,此方法能够提高图像中灰度分布不均匀区域的目标分割精度,加快了收敛速度,并能准确定位目标对象的轮廓位置。
CV model based on global information can not effectively segment the image with uneven gray level, however image local information can better reflect the image characteristics of the target.Based on the Chan-Vese model, the local information of the target is incorporated into the model, and the relative entropy is used as a measure to minimize the similarity between the inner and outer regions of the contour, so as to improve the accuracy of image segmentation and suppress image noise.Experimental results show that the proposed method can improve the segmentation accuracy of the region with uneven gray distribution in the image.It can not only improve the accuracy of image segmentation, speed up the convergence rate, but also accurately locate the contour of the object.
作者
邓丹君
倪波
DENG Danjun;NI Bo(School of Computer,Hubei Polytechnic University,Hubei Huangshi 435003)
出处
《湖北理工学院学报》
2019年第4期33-37,共5页
Journal of Hubei Polytechnic University
基金
湖北省教育厅科学研究项目(项目编号:B2018247
B2018251)
湖北理工学院科研项目(项目编号:18xjz04C)
关键词
图像分割
活动轮廓模型
全局和局部信息
区域相似度
相对熵
image segmentation
active contour model
global and local information
region similarity
relative entropy