多源遥感数据测绘应用关键技术分析

Analysis of Key Technologies in Mapping and Mapping Application of Multi-source Remote Sensing Data

本文首先概述了基于纹理特征的半自动提取技术,并使用基于Gabor变换的区域生长法从纹理复杂的居民地中勾勒出地物轮廓,总体精度达到了96.4%。随后介绍了几种图像配准方法,并在图像配准的基础上检测地物要素的变化情况,使得检测精度达到了99.45%,该检测结果可直接用于地理国情监测等应用。

This paper firstly outlines the semi-automatic extraction technology based on texture features,and uses the region growing method based on Gabor transform to outline the contours of ground objects from complex textured residential areas,with an overall accuracy of 96.4%.Then,several image registration methods are introduced,and the changes of ground features are detected on the basis of image registration,which makes the detection accuracy reach 99.45%.The detection results can be directly used in applications such as geographic monitoring.