基于PCA光谱增强与区域扩散阈值确定的遥感影像变化检测算法  

Remote Sensing Image Change Detection Algorithm Based on PCA-Enhancement and Regional Diffusion for Threshold Determination

作　　者：艾思敏 邓小炼[1] 何卓丽 

机构地区：[1]三峡大学理学院,湖北宜昌443002

出　　处：《三峡大学学报（自然科学版）》2014年第1期90-93,共4页Journal of China Three Gorges University：Natural Sciences

摘　　要：变化检测技术是从观测数据中有效地提取各种变化信息.本文针对变化检测中差异影像的构造和变化阈值的确定展开研究,首先将两期的遥感影像矢量叠加,然后进行主成分分析获得变化量,最后通过区域扩散阈值确定方法获得变化检测结果.通过与传统变化检测方法的实验比较,证明了该方法的有效性.Change detection can obtain useful changed information from the observed data. In this paper, the change detection research is based on building difference image and threshold determination. Firstly, Two pe- riods of remote sensing image are superimposed~ and then principal component analysis is used to obtain change vector. Finally, the change detection results are obtained through regional diffusion for threshold de- termination. Compared with the traditional method, the proposed algorithm is effective.

关 键 词：变化检测 矢量叠加 PCA 区域扩散 

分 类 号：TP7[自动化与计算机技术—检测技术与自动化装置]