Study On Surface Coverage Information Extraction Based On Object-oriented In Ionic Rare Earth Mining Area

Ionic rare earth is an important strategic resource in the 21 st century.There are many mines in southern Jiangxi.In the process of ion-type rare earth mining,monitoring is costly and difficult,and it is difficult to monitor the reclamation situation and assess the environment of the mining area in the process of environmental recovery after mining.The study of rapid,accurate,real-time and efficient extraction method of earth cover information in rare earth mining area is the key to monitor the reclamation situation and evaluate the mining environment.At present,the surface cover information extraction of rare earth mining areas is mostly confined to single surface feature extraction,which cannot meet the demand of surface cover information extraction of rare earth mining areas.In the process of extracting the surface cover information of rare earth mining area with high-resolution image,the image data is large and the research efficiency is low.In view of the above problems,this paper has carried out the mining area surface cover information extraction and mining area scale effect research,which is of great significance to mine reclamation and restoration and mine environmental monitoring.This paper constructs an object-oriented hierarchical classification method for mining area surface cover information extraction,and carries out multi-scale surface cover information extraction and change monitoring analysis around this method.In the study area,the rare earth mining areas of xunwu county and dingnan county in southern jiangxi province were selected.Based on the characteristics of sedimentation ponds in the process of rare earth mining and the characteristics of tailings land and reclaimed land after ecological reclamation,the object oriented method for ground object recognition of remote sensing image of surface cover in the rare earth mining area was studied.The specific experimental conditions are as follows:1.Using the Pleiades-1A image,the multi-resolution image was obtained by resampling.For remote sensing images of each resolution,the classification results of land cover information were obtained by object-oriented hierarchical classification method and process.Finally,the classification accuracy of ground cover information in the mining area under multi-resolution is calculated,and a map is drawn to summarize and analyze the variation of the extraction accuracy of ground cover information with the resolution,and then the variation law of the ground cover scale in the mining area is studied to find out the resolution selection interval of the image in the mining area.The results show that the image with a resolution higher than 3.5m is needed in the study of the mining area surface cover,and the image with a resolution higher than 2m is needed in the study of the change of the sedimentation tank during mining.2.Based on Pleiades 1A image,GF-2 image and GF-1 image,the optimal segmentation scale corresponding to the surface features in the mining area was calculated.After segmentation calculation,stratified classification was carried out,and classification accuracy was obtained through confusion matrix.The results show that,in the remote sensing images with a resolution of 0.5m-2m after fusion,the object-oriented layered classification method can extract the land cover information with high precision,which is of great popularization value.3.On the basis of Pleiades-1A image and GF-1 image,the mining dynamic monitoring of rare earth mining area was carried out by first classification and then processing method.The changes of rare earth mining area during 2013-2015 were obtained,and the changes of surface cover after the reclamation of rare earth mining area were studied.The experiment shows that the multi-source image can realize the monitoring of the land cover change in the mining area by object-oriented stratified classification method.After the reclamation of the rare earth mining area,the land cover area of the mining area increases significantly,and good results have been achieved in the ecological reclamation in recent years.