摘要
对TM图像作比值及主成分分析 ,选择能突出金矿蚀变的高特征向量值所在的主分量进行彩色合成 ,通过空间滤波及非监督分类处理 ,图像上显示出似晕圈状的彩色异常区。经野外验证 ,部分彩色异常区与化探异常分布范围相叠合 ,新发现的 3处金矿点金品位分别达到 2 .74× 10 - 6 、9.6 1× 10 - 6 和 9.6× 10 - 6 。结合图像上的彩色异常信息和取样分析结果 ,在进行成矿条件分析后 ,提出了三处金 (银铜 )遥感综合成矿预测区。
The methods of TM bands ratio and principle component analysis were used to select the characteristic vector that can stress gold alteration mineral information. Based on the characteristic vector, the color composite images were produced. The space filtering and non-supervised classification methods were further used to process the image. There were halo-shaped color abnormal areas in the image. After field verification, some color abnormal areas tally with the geochemical abnormal distribution. In three new-discovered gold ore occurrence, the tenor of gold is 2.74×10 -6 , 9.61×10 -6 and 9.6×10 -6 . According to the color abnormal information in image and the sample analysis result, after analysis of metallogenetic conditions, three gold (silver, copper) mineralization prognostic area were given.
出处
《国土资源遥感》
CSCD
2000年第3期31-37,共7页
Remote Sensing for Land & Resources
基金
"九五"有色遥感技术应用研究重点项目
关键词
主成分分析
遥感信息
金矿
成矿预测
Principle component analysis
Remote sensing information
Gold deposit
Xiaojiayingzi area
