面向对象与三维可视化方法结合的铁路线选址  

Railway Location Based on Combination of Object-oriented Classification and 3D Visualization

作　　者：王萍[1] 许春梅 WANG Ping;XU Chunmei(School of Surveying,Mapping and Spatial Information,Shandong University of Science and Technology,Qingdao,Shandong 266500,China)

机构地区：[1]山东科技大学测绘与空间信息学院,山东青岛266500

出　　处：《遥感信息》2022年第2期1-6,共6页Remote Sensing Information

摘　　要：针对大范围多类型的地质解译效率低、规范性差的问题,提出采用面向对象分类结合三维可视化的方法,对铜吉铁路沿线进行半自动化解译。首先,以高分一号、哨兵二号为主要数据源,通过主分量运算、图像融合等技术增强不同地物特征,选取不同的分割尺度以及特征参数,采用面向对象方法对地层、水文地质等面状信息进行专题提取,其分类精度均大于0.77,提高了解译精度和效率;然后,结合方向滤波增强线性构造,将各类地质要素进行三维可视化分析;最后,选取解译图、地层解译图和植被覆盖度等九个专题要素,通过权重分析进行综合评价,给出沿线地质稳定性综合指数。结果表明,研究区地质状况较好,该方法具有有效性,为铁路线选址提供技术支持。To solve the problem of low efficiency and poor standardization of geological interpretation in large scale and various types,this paper presents a semi-automatic interpretation method combining the object-oriented classification and 3D visualization with the Tongji railway line as an example.With the GaoFen-1(GF-1)and Sentinel-2 as the main data sources,different terrain features are firstly enhanced through the techniques of principal component calculation and image fusion,and the object-oriented method is used for thematic extraction of faceted information such as stratigraphy and hydrogeology by selecting different segmentation scale and feature parameters,with the classification accuracy greater than 0.77,improving the accuracy and efficiency of interpretation.Secondly,the linear structure is enhanced by directional filtering,and all kinds of geological elements are analyzed by 3D visualization.Finally,nine thematic elements,such as drainage interpretation map,stratum interpretation map and vegetation coverage,are selected and the comprehensive index of geological stability along the railway line is given by weight analysis.The results show that the geological condition of the study area is good and the research method is effective,which can provide technical support for the location of railway line.

关 键 词：目视解译 面向对象分类 地质解译 三维可视化 植被覆盖度 

分 类 号：P237[天文地球—摄影测量与遥感]