Study On Identification Method Of Vietnam Opencast Coal Mining Area Based On Satellite Spectral Data

In modern society,demand for coal is increasing,while opencast coal mines is becoming less and the exploration of coal mining is more and more difficult.In Vietnam,the geology of coal mining is very complex,casualties and low efficiency happens during the coal exploration work.Therefore,it is of great significance to study the exploration technology of opencast coal mine area.With the development of science and technology,remote sensing technology is an important result which mainly applied for resource monitoring and investigation.The thesis starts with remote sensing technology and spectrum of coal.Based on the remote sensing theory and the spectral characteristics of coal,the distribution of the opencast coal mines were measured by remote sensing images.First,the remote sensing technology and coal exploration methods were studied and the advantages and disadvantages of each method were analyzed.Subsequently,the extreme learning machine is studied in details,an improved remote sensing technology combined with the limit learning machine is proposed to explore the coal mines.Secondly,the original remote sensing image data was dealt with the use of radiometric calibration and atmospheric correction method to eliminate the interference of atmospheric,light and other factors.Then,remote sensing classification model of coal combined with the improved ELM neural network algorithm is constructed after sampling of the remote sensing data,and simulation results show the effectiveness of the improved ELM algorithm.Finally,the classification model is applied to the remote sensing image of opencast coal mine.At first,the improved ELM algorithm model is used to identify the coal areas in remote sensing image,and the recognition result is compared with the Google Earth image.Then,the recognition result is compared with other classification methods,revealing that the results of the improved ELM algorithm has a higher recognition accuracy than BP neural network,SVM algorithm,M-RFS algorithm and RF algorithm.The classification model is applied to other opencast coal mines,to test the practicability of the method further.