基于树形结构的多相水平集图像分割算法仿真  

Simulation of Multi-phase Level Set Image Segmentation Algorithm Based on Tree Structure

作　　者：陈娥祥 CHEN E-xiang(Fuzhou Fujian Province 350000,China)

机构地区：[1]福建农林大学东方学院

出　　处：《计算机仿真》2019年第6期214-217,共4页Computer Simulation

基　　金：电子商务专业核心课程工学结合改革探索;项目编号:DFJG2018002

摘　　要：传统的多相水平集图像分割算法普遍存在图像分割运行时间较长、成本消耗较大、错误率较高等问题。为了解决上述问题,提出一种基于树形结构的多相水平集图像分割算法。通过改变图像背景,完成多相水平集的重新收敛;根据图像亮度对图像背景进行填充,引入加性算子分裂格式对水平集图像给出统一的数值。以二相水平集分割方法为基础,利用水平函数进行多相水平图像分割。实验结果表明,所提算法图像分割运行时间较短、成本消耗较小、错误率较低。Traditionally,the multi phase level set image segmentation algorithm generally has the problems a-bout long running time,high cost and high error rate.Therefore,an algorithm to segment multi phase level set im age based on tree structure was presented.By changing the background of image,the multi-phase level set was con verged again.According to the brightness of image,the background of image was filled.Moreover,the additive oper ator splitting scheme was used to give a uniform value to the level set image.Based on the two-phase level set seg mentation method,the level function was used to segment the multi-phase level image.Simulation results show that the proposed algorithm has shorter running time of image segmentation,lower cost and lower error rate.

关 键 词：树形结构 多相水平集 图像分割 水平函数 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]