遥感影像信息提取技术的研究与实现

随着遥感技术的发展和遥感应用的深入,遥感影像信息提取方法的研究成为关注的焦点.一般来说遥感影像信息提取包括分类、识别和特征提取.文中主要研究利用MATLAB实现影像中的光谱特征和纹理特征的提取和分类,并结合ERDAS软件对影像进行预处理和分类后处理以及精度评定,最终从精度评定结果来看,两种特征分类的总体精度较高,另外,图像的结构信息有助于提高遥感影像信息提取的精度.

Research and realization of remote sensing image information extraction technologies

With the development of remote sensing and the deep application of remeote sensing, remote sensing image information extraction method has become the research focus. In general, it includes the classification, identification and feature extraction. It mainly studies the extraction and classification of images＇ spectral and texture feature by using MATLAB software, and combined with ERDAS software,it carries out preparation and post-processing for images and accuracy assessment. In addition, it accomplishes the vegetation extraction of remote sensing images through ERDAS. From the view of accuracy assesssment result, the total accuracy of classification based on spectral and texture feature are both high, so a conclusion is obtained that the spatial structure information is helpful to improve the accuracy of remote sensing image information extraction.