摘要: |
提出一种多分类器融合的快速高维特征聚类图像分割方法,将图像高维
特征数据的分类分解为基于灰度(颜色)特征的最佳模糊分类以及基于空域约束的统计分类等
多个低维特征数据的分类。通过多分类器融合的方法将不同分类器得到的分类结果进行优化
整合,得到最后的分类结果。实验证明:与其它图像分类算法相比,该方法拥有更好的分
割性能并大大提高了计算速度,最大限度地保证了分割算法计算的简单有效性。 |
关键词: 图像分割 高维特征聚类 多分类器融合 |
DOI:10.3969/j.issn.1001-893X. |
|
基金项目:国家自然科学基金重点资助项目(60736045) |
|
Image Segmentation Based on Fast High Dimensional Characteristic Clustering Using Combination of Classifiers |
HUANG Rong-shun,WU Hong-gang,LIU Si-yuan |
(The Second Research Institute of CAAC, Chengdu 610041, China;Institute
of Chengdu, ZTE Corporation, Chengdu 610041, China) |
Abstract: |
A new image segmentation algorithm is proposed which is based on fast hi
gh dimensional characteristic clustering using combination of classifiers.In t
he algorithm, the clustering of high dimensional characteristic data
is divided into optimal fuzzy classifying of grayscale (color) and statistical c
lassifying of spatial constraint information. The classification results of the
two different classifiers are integrated to obtain the final image segmentation
result using combination of classifiers.Experiment result proves the good p
erformance and computation simplicity of the algorithm. |
Key words: image segmentation high dimensional characteristic clustering combination of classifiers |