| 基于分割的纹理分类农田信息提取 |
| |
| 引用本文： | 孙小芳.基于分割的纹理分类农田信息提取[J].闽江学院学报,2010,31(2):132-135. |
| |
| 作者姓名： | 孙小芳 |
| |
| 作者单位： | 闽江学院地理科学系,福建,福州,350108 |
| |
| 基金项目： | 福建省教育厅项目(JA09194);;福建省科技厅项目(2008F5044) |
| |
| 摘 要： | 探讨利用分割和纹理分类技术提取农田信息．利用多分辨率分割快鸟影像4个波段，再对分割所生成的影像目标进行纹理分类．利用自相关函数确定纹理窗口的大小为5×5像元，分类所采用的纹理包括同质度、对比度、不相似度．结果表明该方法分类出的6种农田地物正确分类率在0．918到0．852之间．
|
| 关 键 词： | 多分辨率分割 自相关函数 纹理 |
Segmentation-based texture classification farmland information extraction |
| |
| SUN Xiao-fang.Segmentation-based texture classification farmland information extraction[J].Journal of Minjiang University,2010,31(2):132-135. |
| |
| Authors: | SUN Xiao-fang |
| |
| Institution: | Department of Geography;Minjiang University;Fuzhou;Fujian 350108;China |
| |
| Abstract: | This paper presents segmentation-based texture classification and extraction farmland information. Applicability multi-resolution segmentation four bands of quickbird image, then texture classification image objects that after segmentation. Use of the autoeorrelation function of determining the size of the window texture 5 × 5 pixels, texture knowledge include homogeneity, contrast, dissimilarity. The results show that the classification method of six farmland features in the correct classification rate of between 0.918 to 0. 852. |
| |
| Keywords: | multi-resolution segmentation autocorrelation function texture |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
