Research And Demonstration Application On Interpretation Sign System And Information Extraction Method In Open-pit Mine

China is a developing country with a large population.Meanwhile,as the world's major mining country,various geological environmental issuses caused by long-term extensive mine production and development activities threaten the safety of people's lives and properties,seriously affecting and restricting the socio-economic and mining industries' sustainable development.As an important part of the geological environment,open-pit mining mainly covers the main contents of the mining area's geological background,natural environment,ecological & humanistic environment,and geological disasters.With the help of high-speed development of remote sensing technology,exploring and studying the issues about how to use remote sensing technology in the studying of open-pit mining geological environment,can ease the demand for mineral resources by rapid economic and social development,reduce the geological environment damage caused by the production and development of mineral resources,strengthen the geological environmental protenction and comprehensive managenment of mines,and promote the rational development and utilization of mineral resources,ensure that environment is protected while meeting the requirement of mineral resources.Henan Province is a major province of China's mineral resources reserve,development and utilization.The province has a wide range of mineral resources and complete types of mines.The unique and convenient traffic environment and geographical location determine the active development of mineral resources in Henan Province,which is an ideal research area to study how to use remote sensing technology in geological environment studying.The research work of this paper is based on the practice and application of the remote sensing survey of mine geological environment in Henan Province.This paper chooses Henan Province as the main research area,aim to study how to use remote sensing technology in geological environment.Baed on the relevant theory and research methods of mine geological environment and remote sensing technology,this paper built the remote sensing interpretation mark system of open-fit geological environment in Henan Province,raised up a general working method for remote sensing interpretation of open-fit geological environment,and study the multi-source data collaborative information extraction method basing on object-oriented highresolution remote sensing image information extraction technology.At the same time,two typical demonstration areas of non-metallic research area(in the middle of Henan Province)and metallic mineral research area(in the west of Henan Province)were selected to explore the study on evaluation of open-fit geological environment.The paper is divided into six chapters,the main content and chapters are follows: The first chapter introduces the back ground of the dissertation,including three study area——high-resolution monitoring method of mine geological environment,object-oriented high-resolution satellite image information extraction method and evaluation of mine geological environment effect,this chapter summarize relevant literature,current research results,existing problems and future development trends.The second chapter introduces the traffic location meteorology,hydrology,topography,geological background and development & utilization of mineral reshources in Henan Province;Besides,this chapter introduces the basic situation of two research(non-metallic research area in the middle of Henan Province and metallic mineral research area in the west of Henan Province).The third chapter is based on the concept,research content and major issues of the mine geological environment,summarized and established the basic theory and main research methods for remote sensing research of open-pit geological environment.The fourth chapter is based on the practice and application of the remote sensing survey of mine geological environment in Henan Province;According to different types of minerals and different mining methods,this chapter established a remote sensing interpretation system for open-pit geological environments in Henan Province,with the field of identification signs and the remote sensing interpretation signs;Futhermore,this chapter also proposed a general interpretation woring method.The fifth chapter focuses on the research of multisource data collaborative information extraction method based on object-oriented highresolution remote sensing image information extraction technology applied in the two typical demonstration areas for the research area.The sixth chapter is the conclusion a prospect of the thesis.Chapters 3 to 5 belong to the author's independent research.The main research academic achievements and innovations are as follows:1.The theoretical basis,research contents,and research methods for the remote sensing research of the open-pit geological environment are preliminarily established.According to the existing theoretical basis,research content and main problems of mine geological environment,referring to the scope of remote sensing technology,this article summarizes the concept,research objectives and main classification how to use remote sensing technology in open-fit mine geological environment research.This article combines the remote sensing characteristics of open-pit geological environment and the main research methods of remote sensing technology,preliminarily establishes the openpit geological environment remote sensing research methods,including human-computer interaction visual interpretation,remote sensing image information extraction,objectoriented remote sensing image information extraction method,and multi-source data collaborating target extraction method.2.This article is based on the practice and application of the remote sensing survey of mine geological environment in Henan Province.Using domestic high resolution remote sensing satellite imagery,this article summarizes the basic theories and research content of remote sensing research on open-pit geological environment and the classification of major issues.With the two aspects of field identification signs and remote sensing interpretation signs,this paper proposes a geological environment interpretation sign system for open-pit mines in Henan Province,and clarifies the classification,definition and feature description of various geological environmental issues that arising by different mineral species and different mining methods.3.This article proposes two different methods for extracting multisource data collaborative information based on object-oriented technology:(1)Experiment 1 uses the ESP(Estimation of scale parameters)to calculate local variance of overall image,and determine opimal segmentation parameters by local variance variation.(2)Based on the image information hierarchy,experiment 2 uses the Euclidean two-index method to calculate the optimal segmentation parameters of each layers,and selects an appropriate feature factor to establish a fuzzy rule set;Besides,this experiment uses adopts a layered mask method to separate out the large-scale geological environment background elements;And then,this experiment uses random forest model to classify the rest of the geological environmental categories in metallic mineral research area that located in the west of Henan Province;Finally,with the confusion matrix and Kappa coefficient for accuracy evaluation,the result shows that the classification accuracy(87%)and Kappa coefficient(0.74)of the random forest model were higher than those of the classification and regression tree model(classification accuracy of 79%,Kappa coefficient of 0.6)and the rotation forest model(classification accuracy of 76%,Kappa coefficient of 0.53)in the experiment 1;Besides,the result alos shows that the random forest model has improved the indetification ability of the major open-pit geological environmental factors(classification accuracy is 88%,Kappa coefficient is 0.78)through hierarchical segmentation and extraction in the experiment 2.Through the multi-source data collaboration to eliminate the classification logic error,the two experimental results show that the two methods can effectively extract large-scale major geological environment elements,but the identification ability of the elements with indistinct image features in not enough.