摘要
依据植物图像中不同目标的区域特征,应用多水平集分割算法分割植物图像.该算法能够将植物的花朵、叶片以及背景有效地分割开.与基于聚类的多尺度Ncut算法的分割效果进行比较,多水平集方法在分割效果上优于多尺度Ncut算法.
Based on the region properties of different objects in plant images, segmentation algorithm based on multi-level sets was applied to plant image. The experimental results showed that the algorithm could segment the flower, leaves and background in plant images succesfully. The result suggested that multi-level set algorithm was better than multi-scale Neat algorithm.
出处
《福建农林大学学报（自然科学版）》
CSCD
北大核心
2009年第3期310-313,共4页
Journal of Fujian Agriculture and Forestry University:Natural Science Edition
基金
福建省自然科学基金(2006J0299)
福建省教育厅(2007F5009、JA07077)资助项目
关键词
图像分割
多水平集
植物
image segmentation
multi-level set
plant
