个旧高松矿田线性构造对比分析研究

在遥感影像中,一些断层和断裂受侵蚀作用会显现出线形特征,即线性构造。本文对个旧矿区高松矿田区域遥感影像经降噪处理后采用线段追踪法(STA:Segment Tracing Algorithm)提取线性构造并进行线素连接,连接后的影像进行解译并做数据提取,最终将提取出的对比数据与实测断裂数据进行再对比和相关性分析,分析结果表明:断裂长度和走向的Pearson相关系数分别为0.67和0.987,双尾显著性值分别为0.048和0,说明线段追踪法提取的断层长度和走向与实测数据相关性较高,可以较好地解译出个旧矿区断裂的分布情况;同时,还可以看出部分大断裂交汇处的小断裂发育良好,可为圈定预测成矿靶区提供参考。但是,矿区影像东北区的线性构造分布较少,个松断裂的长度连贯性和芦塘坝断裂的走向都没有得到很好的解译,这可能与图像降噪不完善和线性构造提取中的程序参数设定不准确以及线素连接过程中线素之间的联系和转折过大有关,这些问题有待于在下一步的工作中逐步完善。

Correlation analysis of lineaments of remote sensing images and the measured fractures and faults in Gaosong ore field, Gejiu city, Yunnan province

Faults and fractures in remote sensing image take form of lineaments due to erosion. Denoising procession is made for remote sensing images of Gaosong ore field. STA(Segment Tracing Algorithm) is adopted to extract linear structure and is connected the line elements then the image interpreted and data extracted. Data extracted are compared with those of the measured faults and fractures and their correlation analyzed. Pearson correlation coefficient of length and strike of fault is 0.67 and 0.987, double tail significance value 0.048 and 0 respectively. The fault length and strike extracted by STA are highly correlated with those of the measured fault. The remote sensing image interpretation reflects real distribution of faults and fractures in the Gejiu mining district. In addition small faults or fracturs at cross point of big faults are interpreted supplying reference for locating further prospecting target. However in image of the Northeast mining district less lineaments occur and length and continuation of Gaosong fault and strike of Lutangba fault are not well interpreted, maybe due to imperfect denoising of the image, inaccurate program parameter set for linear structure extraction and exssesive connection and transition between line elements. In future our work should be focused on solving the problem.