基于面向对象分类法的川藏铁路沿线大型滑坡遥感解译

发布时间：2018-04-02 12:16

本文选题：青藏高原　切入点：面向对象分类　出处：《现代地质》2017年05期

【摘要】：川藏铁路位于青藏高原中东部,其地形地貌和地质构造复杂,是我国大型滑坡发育最为密集的地区,严重影响了我国西南地区的人类生命财产安全和重大工程建设。在面上资料分析的基础上,重点以川藏铁路沿线茶树山滑坡、川藏公路102道班滑坡、八宿怒江滑坡、理塘乱石包高速远程滑坡为例,在ENVI5.1和e Cognition软件平台上,利用高分辨率World View-2以及Landsat遥感卫星数据,结合野外实地调查,采用基于面向对象分类法对滑坡的遥感信息进行分析研究。结果表明,采用面向对象分类法可以提取出关键信息和目标区域,再结合目视解译,能够得到滑坡的细部信息,提高遥感影像滑坡解译的成功率,特别是对川藏铁路沿线等地质条件复杂区域的滑坡调查工作有重要意义。最后,结合灰度共生矩阵(GLCM)和归一化植被指数(NDVI)对古滑坡和新生滑坡的识别进行探讨。基于灰度共生矩阵和植被指数提出了滑坡遥感信息量判别(GVI)模型,并构建模型的质量函数I_(GVI);统计样本的结果显示古滑坡的I_(GVI)值明显低于新生滑坡,表明本研究提出的GVI模型可以为识别古滑坡和新生滑坡提供依据。
[Abstract]:The Sichuan-Tibet Railway is located in the central and eastern part of the Qinghai-Tibet Plateau. Its topography, geomorphology and geological structure are complex, and it is the most densely developed area of large-scale landslides in China. It has seriously affected the safety of human life and property and the construction of major projects in the southwest of China. Based on the analysis of the data on the surface, the emphasis is placed on the landslide in Chashushan along the Sichuan-Tibet Railway, the landslide in 102 Daoban on the Sichuan-Tibet Highway, and the landslide in the Baichu River. On the platform of ENVI5.1 and e Cognition software, the high resolution World View-2 and Landsat remote sensing satellite data are used in the remote landslide in Litang, as an example, combined with field investigation, The remote sensing information of landslide is analyzed based on object oriented classification. The results show that the key information and target area can be extracted by object oriented classification, and the detailed information of landslide can be obtained by visual interpretation. It is of great significance to improve the success rate of landslide interpretation from remote sensing images, especially to the investigation of landslides in areas with complicated geological conditions such as along the Sichuan-Tibet railway line. Combined with GLCM) and NDVI (normalized vegetation index), this paper discusses the identification of ancient landslide and new landslide. Based on the gray level co-occurrence matrix and vegetation index, a GVI model is proposed for the identification of landslide remote sensing information. The results of statistical samples show that the value of I _ T _ GVI of ancient landslide is obviously lower than that of new landslide, which indicates that the GVI model proposed in this study can provide the basis for identifying ancient landslide and new landslide.
【作者单位】：中国地质大学(北京)地球科学与资源学院;中国地质科学院地质力学研究所;国土资源部新构造运动与地质灾害重点实验室;
【基金】：中国地质调查局项目(DD20160271,12120113038000) 国家自然科学基金项目(41402321)
【分类号】：P627;P642.22

【相似文献】