摘要
通过对P5立体像对、GeoEye-1影像的处理与分析,分别进行了数字高程模型的提取和地层的划分,用检查点法和目视检查法对提取的数字高程模型的误差评估结果显示,误差在5m以内,满足产状提取的要求。根据地层划分结果,在不同时代的地层边界处和同时代地层内部选择岩层三角面发育较好、产状相对稳定的区域,选取用于地层产状提取的点,并获取其三维坐标,然后在MATLAB环境下根据三点法或多点拟合法原理及公式编程提取地层产状。野外验证结果表明,提取的产状倾角误差均在6°以内,其中85%的误差在4°以内,中误差在3°以内,均在定量研究可接受的范围之内,并分析和总结了误差产生的主要原因。研究结果表明,利用P5立体像对与GeoEye-1影像提取近地表地层产状的方法,有效地约束了浅层构造的形态,减少了构造分析的不确定性,对柯坪的构造变形研究具有重要意义。
In this paper,P5 stereo images and GeoEye-1 image are firstly processed and analysed in order to extract digital elevation model（DEM）and divide stratum.The perpendicular root mean square of extracted DEM is within 5meters in residual elevation report using methods of check point and visual inspection.According to the stratigraphic classification,measure points which are located at boundary or inner stratum are chosen at regions with better triangular facet of rock and stable attitude,and then the attitude of stratum is extracted using three-dimensional data of the measure points through programming in MATLAB environment based on three-point method or multiple-point fitting method.Based on field survey,the measurement residual of dip angles is within 6 degrees,and 85%are less than 4 degrees.And the root mean square of the angles is less than 3 degree.At the same time,seasons of residual generation are concluded,including geological structure,topography,accuracy of DEM,matching degree of DEM and images,the location of chosen point and so on.It is suggested that the method of stratigraphic attitude extraction using P5 stereo images and GeoEye-1 image is less restricted by worse natural conditions.Not only can the method restrict the profile of shallow tectonics,but also can cover the shortage of attitude of stratum and have great significance for reducing uncertainty of Kalpin thrust tectonic deformation research.It provides a new way for quantitative research of structural geology,and also provides references for similar regions.
出处
《地震地质》
EI
CSCD
北大核心
2011年第4期951-962,共12页
Seismology and Geology
基金
国家自然科学基金(40872130)
财政部重大专项“我国地震重点监测防御区活动断层地震危险性评价”共同资助
