| 基于树型增长神经网络模型的遥感图像聚类 |
| |
| 引用本文： | 刘文岭,郑小慎. 基于树型增长神经网络模型的遥感图像聚类[J]. 天津科技大学学报, 2006, 21(1): 44-46,68 |
| |
| 作者姓名： | 刘文岭 郑小慎 |
| |
| 作者单位： | 天津科技大学海洋科学与工程学院,天津,300457 |
| |
| 基金项目： | 天津市高等学校科技发展基金 |
| |
| 摘 要： | 通过对天津海岸带遥感图像的研究，提出了基于树型增长神经网络模型的遥感图像聚类方法。该方法申神经模型的网络结构在训练过程中动态生成，用户可根据需要实现层次聚类，同时可以通过调节扩展因子SF的大小调节聚类的速度和精度，从而提高了聚类的精度和灵活性。
|
| 关 键 词： | 神经网络 遥感图像 聚类 |
| 文章编号： | 1672-6510（2006）01-0044-03 |
| 收稿时间： | 2005-08-30 |
| 修稿时间： | 2005-08-30 |
Remote Sensing Image Clustering Based on Tree-Structured Growing Self-Organizing Feature Map |
| |
| LIU Wen-ling,ZHENG Xiao-shen. Remote Sensing Image Clustering Based on Tree-Structured Growing Self-Organizing Feature Map[J]. Journal of Tianjin University of Science & Technology, 2006, 21(1): 44-46,68 |
| |
| Authors: | LIU Wen-ling ZHENG Xiao-shen |
| |
| Abstract: | This paper is presented that the novel method of remote sensing image clustering-based on the tree-structured growing self-organizing feature map (TGSOM) , which has a dynamic tree-structure generated during the training process, at the same time, TGSOM 's growth speed can be controlled through the function of the spread factor(SF) , and the precision of clustering results is different because of the difference value of SF. The user can get the hierarchical clustering results through changing the size of SF in different steps during clustering, which can improve the precision and flexibility of the clustering results. |
| |
| Keywords: | neural network remote sensing image clustering |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
