摘要
多光谱影像具有丰富的地物信息,利用多光谱影像进行聚类分析可以充分利用遥感影像为在大范围地区进行变化检测和土地利用分析作出应有的贡献。而模糊聚类更接近实际情况,能提高影像分类精度,缩短分类时间,提高系统运行效率。本文采用对多光谱影像先进行K-L变换去除了大量的冗余数据后,再进行改进的加权模糊聚类处理方法,取得了较理想的分类效果。
Multi-spectral image has abundant information. Using multi-spectral image, clustering analysis can make full use of the remote-sensing image contributing to the change detection and land use analysis in large zones. The Fuzzy Clustering is more suiting with the real condition; it can reduce the clustering time, and improve the efficiency of the clustering system. In this paper, K-L transformation is used to wipe off the flock redundant data which is contained in the multi-spectral image. Then the improved Fuzzy Clustering is used to achieve a good clustering result.
出处
《地理空间信息》
2006年第6期50-52,共3页
Geospatial Information
基金
国家重点基础研究发展计划(973计划)资助项目(2006CB701303)。
关键词
多光谱
模糊聚类
K—L变换
影像分类
multi-spectrum
fuzzy clustering
K-L transformation
image classification
