| 基于SVM的多源遥感影像分类研究 |
| |
| 引用本文： | 贾萍,李海涛,林卉,顾海燕,韩颜顺.基于SVM的多源遥感影像分类研究[J].测绘科学,2008,33(4). |
| |
| 作者姓名： | 贾萍 李海涛 林卉 顾海燕 韩颜顺 |
| |
| 作者单位： | 国土资源部信息中心,北京,100812;中国测绘科学研究院摄影测量与遥感研究所,北京,100039;徐州师范大学国土信息与测绘工程系,江苏,徐州221116 |
| |
| 基金项目： | 国家高技术研究发展计划 |
| |
| 摘 要： | 本文通过分析单源遥感影像分类的现状和困难,以SAR和SPOT-5影像为实验数据,提出了基于支持向量机(Support Vector Machine,SVM)理论的多源遥感影像分类方法。研究结果表明,本文的方法能够有效地解决单源影像信息分类效果破碎的问题,正确识别地物,对高维输入向量具有高的推广能力,正确率达到94.97%,比多源影像的最大似然分类(Maximum Likelihood Classification,MLC)方法正确率更高。
|
| 关 键 词： | 支持向量机 多源影像 最大似然分类 精度评价 |
Research on multi-source remote sensing image classification based on SVM |
| |
| JIA Ping,LI Hai-tao,LIN Hui,GU Hai-yan,HAN Yan-shun.Research on multi-source remote sensing image classification based on SVM[J].Science of Surveying and Mapping,2008,33(4). |
| |
| Authors: | JIA Ping LI Hai-tao LIN Hui GU Hai-yan HAN Yan-shun |
| |
| Abstract: | On the basis of analyzing the actuality and difficulty of single-source image classification,a method of applying Support Vector Machine to multi-source Remote Sensing image classification is presented in this paper where SAR and SPOT-5 data are used.The results show that this method can solve the image classification fragmentation which is based on the single-source data,recognize objects accurately,has the good generalization ability with the high dimension vector,and has total accuracy of about 94.97%.It has more accuracy than the Maximum Likelihood Classification(MLC) method using multi-source data. |
| |
| Keywords: | support vector machine multi-source image maximum likelihood classification accuracy evaluation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
