神经网络法在寻找"隐伏矿(盲矿)"中的应用机理探讨  

NEURAL NETWORKS CLASSIFIED APPLICATION IN "INSIDIOUS OR SCOTOMA MINE"ITEM

作　　者：张维宸[1] 刘建芬 

机构地区：[1]河北省遥感中心,石家庄050021 [2]河北省地质调查院,石家庄050051

出　　处：《地质与勘探》2003年第2期56-60,共5页Geology and Exploration

基　　金：中国地质调查局科研项目(编号:200020100040696;科研编号:DK0002017)资助。

摘　　要：随着计算机应用技术的不断提高,神经网络作为一个数据处理结构应运而生。传统的遥感找矿方法是通过目视解译圈定"靶区",但在"隐伏矿(盲矿)找矿"项目中却遇到了难题,很难找到相关弱信息,利用神经网络的模拟人脑功能来提取矿化相关弱信息寻找"隐伏矿(盲矿)"是可能的。在此仅仅对其应用做简单的介绍,希望对今后的遥感找矿产生一定的提示作用。With continuous development of computer technology, neural networks as a data-processing structure emerge as the times require. Traditional remote sensing to seek ores is depended on eye observation/explanation, and translation to mark with circle 'target area', but it meets a difficult problem in 'hidden and blind ores'. To find interrelated weak information is very difficult. Neural networks can simulate human brain function to pick-up interrelated mineralization weak information in search for 'hidden and blind ores'. Application of neural networks is simply introduced, and hoped to point out to find ores using remote sensing technology in the future.

关 键 词：神经网络法 遥感 找矿方法 隐伏矿床 

分 类 号：P612[天文地球—矿床学] P627[天文地球—地质学]