铁路工程地质遥感图像处理系统设计与实现  

Design and Implementation of Remote Sensing Image Processing System for Railway Engineering Geological Survey

作　　者：刘桂卫 全德武 

机构地区：[1]中国铁路设计集团有限公司,天津300142 [2]兰州交通大学,兰州730070

出　　处：《铁道工程学报》2017年第8期11-15,52,共6页Journal of Railway Engineering Society

基　　金：中国铁路总公司科技开发重点课题资助(2015G002-0);中国铁设重点课题(721742)

摘　　要：研究目的:铁路行业尚缺乏专业的工程地质遥感图像处理系统,为满足铁路工程地质遥感判释工作需求,需开发具备最佳波段组合方式选择、波段运算、图像融合定量评价、图像增强运算及专题信息提取等功能的软件系统。研究结论:(1)最佳指数法能够辅助波段组合方式的选择;(2)基于光谱逼真度和纹理清晰度指标对图像融合结果进行定量评价,能够优选出最佳融合算法,提高了影像的可判释能力;(3)去条纹功能消除了传感器标定差异,增强了影像清晰度;(4)水体指数法适用于提取平坦区域水体信息,在细节反映上较谱间信息法更好;谱间信息法能将水体与阴影区分开来,适合陡峻山区水体的提取;(5)该软件系统能为铁路工程地质遥感图像处理提供基础支持,可广泛应用于工程地质遥感判释前期数据处理工作。Research purposes： There has not been a professional software of remote sensing image processing for railway engineering geological survey. In order to fulfill the requirements of railway remote sensing interpretation work, we built a software which can realize the optimal bands composite selection, band operation, image fusion and quantitative evaluation, image enhancement and thematic information extraction, etc. Research conclusions： （1）The optimum index method can realize the bands composite selection. （2）Based on the quantitative evaluation through spectral fidelity and texture clarity index, the preferred data fusion methods can be identified, which can improve the interpretation ability of the image. （3） The stripe noise removal method can eliminate the sensor calibration differences, that enhance the image clarity. （4）The water index method, which is better than the spectro -photometric method on the details reflect, is suit for extracting water information in flat areas, otherwise, the spectro - photometric method, which can distinguish the water and shade, is suit for extracting water information in steep mountainous areas. （5）The software provides the basic support for the remote sensing image processing of railway engineering geology survey, and can be widely applied to the remote sensing interpretation of engineering geology work.

关 键 词：地质遥感 融合评价 图像增强 专题信息提取 软件系统 

分 类 号：P642.4[天文地球—工程地质学]