基于贝叶斯分类器的图像分类.pdfVIP

4 5 2 ( ) V ol. 4 5 N o. 2 2 007 3 JOURNAL OF JILIN UN IV ERSITY ( SC IEN CE ED ITION ) M ar 2 007 1 1 2 3 ( 1. 130012; 2. 130022; 3. 130022) : 通过对朴素贝叶斯分类器的讨论 提出将贝叶斯方法应用于医学图像分割后的图像分 类思想. 给出 种基于朴素贝叶斯分类器的图像分类方法 对从尿沉渣图像中识别出的微粒 进行正确分割及特征提取与选择 并利用朴素贝叶斯分类器进行分类. 实验结果表明 所提 出的方法用于解决图像分类有效. : 贝叶斯分类器; 图像分类; 特征提取; 特征选择 : TP 18 : A : 167125489( 2007) 0220249205 Im age C lassifica tion Based on Bayesian C lassifier 1 1 2 3 DONG L i2yan YUAN Sen2m iao LIU Gu ang2yu an JIA Shu2hong ( 1. College of Computer Science and Technology, Jilin University, Changchun 1300 12, China; 2. College ofCommunication Engineering, Jilin University, Changchun 130022, China; 3. Department ofF oundation, viation University of ir Force, Changchun 130022, China) Ab strac t: N a ive Bayesian classifier was discu ssed. N aive B ayesian approach was u sed in mi age classification. Th e n ew m ethod mi proves th e accuracy of mi age recogn ition. An mi age classificat ion m ethod based on na ive Bayesian c lassifier was introdu ced. A fter d istingu ish ing th e v isib le composition s feature extraction and feature selection were processed. In th e end na ive Bayesian c lassifier was u sed in classify ing of object ent ities. Exp ermi ent resu lt sh ow s that th e n ew m ethod is su itab le for mi age c lassification. Ke y wo rd s: Bayesian classifier; mi age classificat ion; feature extraction; feature se lection [ 1] . .