基于极限学习机的遥感地球化学反演模型  

Remote Sensing Geochemical Inversion Model by Using Extreme Learning Machine

作　　者：孙立影 杨晨[1] 赵海士 常志勇 Sun Liying;Yang Chen;Zhao Haishi;Chang Zhiyong(College of Earth Sciences,Jilin University,Changchun 130061,China;College of Computer Science and Technology,Jinlin University,Changchun 130012,China;College of Biological and Agricultural Engineering,Jilin University/Key Laboratory of Bionic Engineering(Jilin University),Ministry of Education,Changchun 130022,China;National-Local Joint Engineering Laboratory of In-Situ Conversion,Drilling and Exploitation Technology for Oil Shale,Jilin University,Changchun 130021,China)

机构地区：[1]吉林大学地球科学学院,长春130061 [2]吉林大学计算机科学与技术学院,长春130012 [3]吉林大学生物与农业工程学院/工程仿生教育部重点实验室(吉林大学),长春130022 [4]吉林大学油页岩地下原位转化与钻采技术国家地方联合工程实验室,长春130021

出　　处：《吉林大学学报（地球科学版）》2020年第6期1929-1938,共10页Journal of Jilin University：Earth Science Edition

基　　金：国家自然科学基金项目(61572228);吉林省科技发展计划项目(20190303006SF,20190302107GX);吉林省产业技术研究与开发专项(2019C053-5,2019C053-7)。

摘　　要：地球化学勘查研究涉及大量采样工作,但在工作环境恶劣人们难以到达的地区,大范围、大比例尺的地球化学数据极难获取。本文基于极限学习机(ELM)构建遥感地球化学反演模型,弥补因为区域数据不足导致的找矿工作困难。依据偏最小二乘回归(PLSR)方法选取与地球化学数据相关性强的遥感影像成分,并根据极限学习机建立地球化学数据与遥感影像之间的非线性对应关系来获取未知地球化学异常,以此来指导找矿工作。实验中,选取研究区铜元素1∶20万土壤地球化学数据与Landsat 8 OLI遥感影像进行反演分析。实验结果表明,基于ELM的遥感地球化学反演所取得的异常分布与已知矿点具有很好的对应度,验证了本文所提出模型的有效性。Geochemical exploration research involves a large amount of sampling work, which is extremely difficult in inaccessible terrain with harsh working environments. The authors propose a geochemical inversion model with remote sensing images by using extreme learning machine(ELM) to alleviate the difficulty of ore prospecting in the areas with insufficient regional data. The partial least squares regression(PLSR) method is used to select the remote sensing image features which are highly correlated with geochemistry data. In this model, the nonlinear relationship between the geochemical data and the remote sensing images is established using ELM for getting unknown geochemical anomalies, after which the ore prospecting work can be further promoted. In the experiment, 1∶200 000 soil geochemical data of Cu element and the Landsat 8 OLI remote sensing images were used for the inversion analysis. The experimental results showed that the anomalous distribution obtained by the ELM-based inversion model had a good correspondence with known ore spots, which verified the effectiveness of the proposed model.

关 键 词：极限学习机 偏最小二乘回归 遥感地球化学 反演 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]