淮北师范大学学报（自然科学版）
淮北師範大學學報（自然科學版）
회북사범대학학보（자연과학판）
JOURNAL OF HUAIBEI COAL INDUSTRY TEACHERS COLLEGE(NATURAL SCIENCE)
2014年
2期
66-70
,共5页
IKONOS影像%分类%支持向量机%小波核
IKONOS影像%分類%支持嚮量機%小波覈
IKONOS영상%분류%지지향량궤%소파핵
IKONOS images%classification%support vector machine%wavelet kernel
针对IKONOS影像波段和光谱特点，结合小波函数对非线性信号的良好描述能力和多输出支持向量回归（MSVR）在多维机器学习领域的优势，提出一种基于小波核MSVR的IKONOS影像多分类方法，并以安徽大学新校区的IKONOS影像进行仿真实验。结果表明，提出的多分类算法优于传统的有监督二分类方法和无监督K-Means分类方法，获得较好的分类效果。
針對IKONOS影像波段和光譜特點，結閤小波函數對非線性信號的良好描述能力和多輸齣支持嚮量迴歸（MSVR）在多維機器學習領域的優勢，提齣一種基于小波覈MSVR的IKONOS影像多分類方法，併以安徽大學新校區的IKONOS影像進行倣真實驗。結果錶明，提齣的多分類算法優于傳統的有鑑督二分類方法和無鑑督K-Means分類方法，穫得較好的分類效果。
침대IKONOS영상파단화광보특점，결합소파함수대비선성신호적량호묘술능력화다수출지지향량회귀（MSVR）재다유궤기학습영역적우세，제출일충기우소파핵MSVR적IKONOS영상다분류방법，병이안휘대학신교구적IKONOS영상진행방진실험。결과표명，제출적다분류산법우우전통적유감독이분류방법화무감독K-Means분류방법，획득교호적분류효과。
According to spectral and band characteristics of IKONOS images,combined with the good descrip-tion ability of wavelet function for nonlinear signal and the advantages of multi-output support vector regres-sion(MSVR)in the field of multi-dimensional machine learning,a multi-classification method is proposed for IKONOS images based on wavelet kernel MSVR. IKONOS images of the new campus of Anhui university are used for experimentation. The experimental results show that the proposed multi-classification algorithm is superior to the traditional supervised two-category method and unsupervised K-means classification meth-od,and obtains better classification results.
