结合图像全局和局部信息的符号压力函数分割模型  

A Signed Pressure Force Function Segmentation Model Combining Global and Local Information of Image

作　　者：孟新超 司智勇[1] MENG Xinchao;SI Zhiyong(Zhengzhou Railway Vocational and Technical College,Zhengzhou 451460,China)

机构地区：[1]郑州铁路职业技术学院,河南郑州451460

出　　处：《郑州铁路职业技术学院学报》2023年第3期31-36,共6页Journal of Zhengzhou Railway Vocational and Technical College

摘　　要：仅采用局部或全局图像信息很难处理弱边界和噪声图像的分割问题,据此提出一种结合图像全局和局部信息的符号压力函数分割模型。首先,利用图像的全局和局部区域信息,分别构造全局和局部灰度拟合项;其次,将定义的全局拟合项与局部拟合项进行线性组合,构造一个混合符号压力函数;再次,重新定义气球力函数,以期达到自适应改变水平集演化速率的目的。数值仿真实验结果表明,该算法不仅能准确分割弱边界和多目标图像,而且分割速度很快且对噪声具有一定的鲁棒性。As we all know,it is difficult to deal with the weak boundary and noisy images by using only local or global image information.Therefore,this paper proposes a signed pressure force function segmentation model combining global and local image information.First,the global and local gray fitting terms are constructed by using the global and local region information of the image respectively.Then,the global and local term are linearly combined to construct a mixed signed pressure force function.Finally,the balloon force function is redefined in order to adaptively change the evolution rate of level set.The numerical simulation results show that the proposed algorithm not only can segment weak boundary and multi-target images accurately,but also has a fast segmentation speed and a certain robustness to noise.

关 键 词：图像分割 活动轮廓模型 水平集 符号压力函数 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]