基于双边滤波的主动轮廓模型细胞图像分割方法研究  

Research on Cell Image Segmentation Method of Initiative Contour Model Based on Bilateral Filtering

作　　者：马竟锋 李晓旭 罗琳[1] 祁鑫[1] 

机构地区：[1]中国石油大学(华东)计算机与通信工程学院,青岛266580

出　　处：《价值工程》2012年第25期194-195,共2页Value Engineering

基　　金：中央高校基本科研业务费专项资金;项目编号:12CX04076A;山东省自然科学基金面上项目;项目编号:ZR2012HM060

摘　　要：区域主动轮廓模型采用先验知识指导建立分割模型,并在分割的过程中采用水平集演化的方式使得零水平集自动收敛于目标物体的边界。在采用该模型进行分割之前,通常都采用高斯滤波器对图像进行滤波,但是高斯滤波器在对物体进行平滑操作的同时,也会造成物体边界的弱化,从而影响分割效果。本文将双边滤波和区域主动轮廓模型结合起来,充分利用双边滤波在平滑目标内部的同时不弱化边界的特点,在利用区域主动轮廓模型进行分割之前使用双边滤波对图像进行处理,实验结果显示使用本方法进行分割得到的结果准确性得到了提高。Regional initiative contour model uses priori knowledge to guide the establishment of segmentation model,and adopts level set evolution to make the zero level set automatic convergence in the boundary of the target object in the process of segmentation.Before conducting segmentation by the model,we adopts Gaussian filter to filter the image,but it also cause a weakening of the object boundary when Gaussian filter conducts the smooth operation on the objects,thus affecting the segmentation results.This paper combines with bilateral filtering and regional active contour model,takes full advantage of the characteristics which the bilateral filter does not weaken the boundary when conducting the smooth operation on the objects,and uses bilateral filter for image processing before segmentation by using regional initiative contour model.Experimental results show that this method can improve accuracy when conducting segmentation.

关 键 词：图像分割 双边滤波 区域主动分割模型 水平集 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]