基于Watershed和Snake混合模型的图像分割算法  

A Hybrid Image Segmentation Algorithm Model Based on Watershed and Snake

作　　者：千堃 马银平[1] 韩悬 QIAN Kun;MA Yin-ping;HAN Xuan(College of Information Engineering,Nanchang Hangkong University,Nanchang 330063)

机构地区：[1]南昌航空大学信息工程学院,南昌330063

出　　处：《现代计算机》2021年第1期62-67,共6页Modern Computer

摘　　要：针对传统Snake模型存在着对初始轮廓曲线位置敏感和不能收敛到凹陷区域的问题,提出将改进的分水岭算法应用于优化Snake模型边缘曲线中。同时再利用遗传算法的全局优化特性,来解决Snake模型的局部极小值问题,提升Snake模型的分割精度。实验结果表明,改进的Snake分割算法不仅能有效解决传统Snake分割算法对初始轮廓位置敏感以及不能收敛到凹陷区域的问题,而且具有分割效率高、分割结果更加精确的优点。In view of the problem that traditional Snake model is sensitive to the position of the initial contour curve and cannot converge to the con⁃cave area,the improved watershed algorithm is proposed to be applied to Snake model.The boundary obtained by the improved watershed algorithm is taken as the initial edge contour of Snake model,and in order to address the problem that the algorithm is easily trapped into lo⁃cal minimum,the results of Snake model segmentation are further optimized by using the global optimization of genetic algorithm.The ex⁃perimental results show that the improved Snake segmentation algorithm can not only effectively solve the problem that traditional Snake model exits,but also has the advantages of high segmentation efficiency and more accurate segmentation results.

关 键 词：分水岭变换 SNAKE模型 轮廓提取 遗传算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]