摘要
分类的目的就是根据现有的图像特征建立一个分类器,能够对未知的图像类型进行预测。在现有众多分类算法中,贝叶斯分类器由于其坚实的数学理论基础并能综合先验信息和数据样本信息,成为当前机器学习和数据挖掘的研究热点之一。本文论述了内容图像检索中基于贝叶斯分类器的图像分类技术。介绍了贝叶斯分类器,叙述了利用贝叶斯分类器进行图像分类的方法,以及图像特征的分布假定。最后通过对分类器的探讨,总结了贝叶斯估计分类的不足。
The aim of classification is to establish a classifier on the basis of the image features, and forecast the unknown image type. Among the lots ofcurrent methods, Bayesian classifier becomes one of the hot problems studied about machine learning and data mining, because of its basis of mathematics and the integrationg of a prior information and data sample information. The technology of image classification on Bayesian classifier was discussed, which based on content-based image retrieval. Bayesian classifier was first introduced, and then the method of image classification using Bayesian classifier and the distributing assumption of image feature were summarized. Finally summarized the lack of Bayesian estimate through the discuss of classifier.
出处
《长春理工大学学报(自然科学版)》
2009年第1期132-134,共3页
Journal of Changchun University of Science and Technology(Natural Science Edition)
关键词
基于内容的图像检索
贝叶斯分类器
图像分类
content-basedimageretrieval(CBIR)
Bayesian classifier
image classification