| 基于多分类器组合面向对象遥感影像信息提取方法 |
| |
| 引用本文： | 王惠英,苏伟,周军其.基于多分类器组合面向对象遥感影像信息提取方法[J].测绘工程,2009,18(5):22-26. |
| |
| 作者姓名： | 王惠英 苏伟 周军其 |
| |
| 作者单位： | 1. 武汉大学遥感信息工程学院,湖北,武汉,430079
2. 中国农业大学信息与电气工程学院,北京,100083 |
| |
| 基金项目： | 国家高技术研究发展(863)计划 |
| |
| 摘 要： | 提出最近距离法和基于知识规则的模糊分类法相结合的混合分类法,针对IKONOS遥感影像,分别用最近距离法、基于知识规则的模糊分类法以及混合分类法对影像进行信息提取。结果表明：混合分类法的信息提取精度最高,总体精度提高到95.60%,Kappa系数提高到0.944,其为面向对象的高分辨率影像信息提供理想方法。
|
| 关 键 词： | 多分类器组合 最近距离法 模糊分类法 多尺度分割 面向对象 |
Object-oriented information extraction method of remote sensing image based on multi-classifier mixed |
| |
| WANG Hui-ying,SU Wei,ZHOU Jun-qi.Object-oriented information extraction method of remote sensing image based on multi-classifier mixed[J].Engineering of Surveying and Mapping,2009,18(5):22-26. |
| |
| Authors: | WANG Hui-ying SU Wei ZHOU Jun-qi |
| |
| Institution: | WANG Hui-ying , SU Wei, ZHOU Jun-qi(1. School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079,China; 2. College of Informa tion and Electrical Engineering, China Agricultural University, Beijing 100083, China) |
| |
| Abstract: | The paper explores a new method which is combined by the nearest neighbor classification method and the fuzzy classification method,and the paper is based on IKONOS remote sensing image,using the nearest neighbor classification method,the fuzzy classification method and the hybrid method to extract information from the image.The result shows： the classification accuracy is the highest by using the hybrid method;the overall accuracy is improved to 95.60%;Kappa coefficient is improved to 0.9440,so it is a satisfactory method of object-oriented to extract information. |
| |
| Keywords: | multi-classifier mixed the nearest neighbor classification method the fuzzy classification method multi-scale segmentation object-oriented |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
