| 基于分形纹理的遥感影像土地覆盖的分类方法研究 |
| |
| 引用本文： | 王娟,张军,吕兆峰.基于分形纹理的遥感影像土地覆盖的分类方法研究[J].测绘科学,2008,33(2):15-17,32. |
| |
| 作者姓名： | 王娟 张军 吕兆峰 |
| |
| 作者单位： | 辽宁工程技术大学,辽宁阜新,123000;中国测绘科学研究院,北京,100039;中国测绘科学研究院,北京,100039 |
| |
| 摘 要： | 提出一种基于分形理论和改进模糊C均值聚类的遥感图像非监督分类方法,该方法尝试将图像的光谱信息和纹理特征相结合。将图像进行主分量变换,根据第一主分量计算图像的布朗运动的各方向的分形维数,差分盒维数和"空隙"等纹理特征作为分类依据。采用改进的模糊C均值聚类,并用混淆矩阵方法评定分类结果精度。通过对试验区的分类试验,说明该方法对改善土地覆盖分类精度行之有效。
|
| 关 键 词： | 纹理 分形 布朗运动 差分盒维数 空隙 模糊C均值 混淆矩阵 |
| 文章编号： | 1009-2307（2008）02-0015-04 |
| 收稿时间： | 2006-11-09 |
| 修稿时间： | 2006年11月9日 |
Study on classification of land cover with remote sensing image based on fractal texture |
| |
| WANG Juan,ZHANG Jun,L Zhao-feng.Study on classification of land cover with remote sensing image based on fractal texture[J].Science of Surveying and Mapping,2008,33(2):15-17,32. |
| |
| Authors: | WANG Juan ZHANG Jun L Zhao-feng |
| |
| Institution: | WANG Juan,ZHANG Jun,L(U) Zhao-feng |
| |
| Abstract: | This paper presents a unsupervised classification method of remote sensing image based on Fractal Theory and ameliorated Fuzzy C-Means Clustering(FCMC), which takes advantages of both spectral and textural features. Firstly, principal-component transformation is performed with the hel Pof several classification elements such as various directions fractal dimensions of Fractal Brown Motion, Difference Box-Counting dimension and lacunarity and so on. Then, with improved FCMC as classifier, the precision of classification results with Confusion-Matrix is assessed. Finally, the experiment of classifying the land cover of remote sensing image has been done and the result is satisfactory, which verifies the effect of this method. |
| |
| Keywords: | texture fractal Brown motion difference box-counting dimension lacunarity fuzzy C-means confusion-matrix |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
