结合局部熵梯度的自适应水平集演化模型  

Adaptive Level Set Evolution Model Combined with Local Entropy Gradient

作　　者：朱维莉 邓小华 王艳 ZHU Weili;DENG Xiaohua;WANG Yan(School of Mathematical Sciences,Chongqing Normal University,Chongqing 401331;College of Training,Sichuan Open University,Chengdu 610073)

机构地区：[1]重庆师范大学数学科学学院,重庆401331 [2]四川开放大学培训学院,成都610073

出　　处：《工程数学学报》2024年第6期991-1005,共15页Chinese Journal of Engineering Mathematics

基　　金：国家自然科学基金(11901071);重庆市自然科学基金(cstc2019jcyj-msxmX0219)。

摘　　要：针对基于图像边缘信息的活动轮廓模型对初始轮廓敏感问题,结合图像局部熵信息,提出一个新的结合局部熵梯度的自适应水平集演化模型。基于图像局部熵梯度和正则化的图像梯度,构造能够随图像信息自适应地改变符号的面积项权重系数,用其代替常用的常值系数,使得演化函数能够根据图像信息自适应上下运动。所提模型从根源上解决初始轮廓敏感问题,允许水平集常值初始化。实验验证了模型对部分灰度不均图像、噪声图像、多目标以及主观边缘图像等均有良好的分割效果。另外,所提的自适应权系数可以用于其他含有面积能量项的分割模型中。To solve the problem that active contour model based on image edge information is sensitive to initial contour,a new adaptive level set evolution model combining local entropy gradient is proposed.The local entropy gradient and regularized image gradient are used to construct the area term weight coefficients to replace the original constant coefficient.The novel weight coefficients can change the symbol adaptively with the image information so that the evolution function adaptively moves up and down according to the image information,allowing constant initialization.The proposed model solves the initial contour sensitivity problem and allows constant initialization of the level set.Experiments verify that the model has good segmentation effect on gray uneven image,noise image,partial multi-object image and subjective edge image.In addition,the proposed adaptive weight coefficients can be used in other segmentation models which involve area energy terms.

关 键 词：活动轮廓 图像局部熵 常值初始化 偏微分方程 

分 类 号：TP391.412[自动化与计算机技术—计算机应用技术]