Extraction Of Coal Mining Collapse And Crack Information Based On Airborne LiDAR Data

In recent years,with the changes of coal mining policy in China,coal mining areas have gradually shifted to the west.In the western ecologically fragile areas represented by the loess gully area,the phenomenon of a large number of ground fissures caused by mining is very common.Compared with the eastern plains,the large number of coal mining cracks in the mining area will cause more serious soil erosion problems in the loess gully area,more likely to cause various geological disasters,and increase the pressure on sustainable land use;traditional methods of manual field surveys coal mining collapse and cracks are monitored,but using traditional methods to monitor coal mining cracks is inefficient,dangerous,and subjective.there are few researches on automatic monitoring of coal mining collapse and cracks,especially the digital extraction of cracks in loess gully areas has not yet formed a set of suitable extraction schemes.This paper took the coal mining subsidence crack area of Ningtiaota coal mine in Shenmu City,Shaanxi Province as the research area.Based on the airborne LiDAR technology and crack extraction algorithm,the ground cracks in the mining area are identified and extracted,and the area,length and width of the cracks in the sample area were further analyzed.Calculated the fracture parameters;sum up the crack information extraction plan in loess gully area with airborne LiDAR as the data source;combine the data of typical crack depth,crack step height and crack width collected manually on the spot to obtain the best crack depth response,according to the crack map of the study area extracted by the optimization algorithm,the mining area is evaluated for the degree of surface fragmentation.The main work and conclusions of this paper are as follows:(1)The selection of DEM interpolation method and the comparison of image filtering.Perform point cloud denoising and point cloud filtering on the airborne LiDAR point cloud.In order to obtain the DEM interpolation method suitable for the two terrains with high precision,the point clouds of the collected two terrains are automatically filtered,and then three interpolation methods are used to generate the DEM on the separated ground points,and the DEM interpolation accuracy is measured by the medium error.Evaluation,and finally got the DEM accuracy of IDW interpolation method to generate two kinds of terrain.The collapsed crack in the mining area is a sudden change of terrain elevation,so the slope map derived from DEM is used as the original image of the crack extraction algorithm.In order to avoid the influence of image noise on the crack extraction effect,and to make the crack information on the slope map more prominent.Three image filtering methods are used to filter the slope map of the sample area.By comparing and analyzing the slope map and the image histogram after image filtering,it is determined that bilateral filtering is the best image filtering algorithm for the experiment.(2)Research on the effect of conventional crack extraction algorithms and optimized crack extraction algorithms on crack extraction.Three common algorithms of threshold segmentation,edge detection,and k-means clustering are used to extract cracks.After comparatively analyzing the advantages and disadvantages of the three crack extraction algorithms,an optimized extraction algorithm combining k-means&Canny was proposed.In order to verify the accuracy of the optimization algorithm in extracting cracks,the recall rate and accuracy rate are used as evaluation indicators to compare the crack extraction accuracy of these several algorithms.The results show that the optimization algorithm has a better effect on the extraction of cracks in the loess gully area.(3)Fracture fine processing and fracture parameter extraction.The mathematical morphology of the crack map extracted by the optimization algorithm is refined,and the morphological principle is mainly used to connect and fill the fracture line and remove the isolated patch of the crack map.The extraction principles of the three fracture parameters of fracture area,length,and width are described,and the three fracture parameters are extracted in the sample area.(4)Establishment of ground damage evaluation and crack depth(visible depth)inversion model in mining area.On the obtained crack binary map of the mining area,the method of dividing statistics is used to calculate the crack area,and then the district area was calculated to obtain the crack density map,and the ground damage degree of the mining area was evaluated according to the crack density range.Use the fracture data collected on the spot to establish the fracture depth inversion model.By comparing and analyzing the fitting degree and average error of various inversion models,the power function of the optimal fracture width and fracture depth is determined as the inversion model of fracture depth.Based on the summary of commonly used crack extraction methods,this paper uses airborne LiDAR as the data source to improve an optimized crack extraction algorithm.Summarized a plan process for the extraction of coal-collapsed crack information suitable for loess gully,and applied experimental results in the research area,evaluating ground crushing extent to the research area and constructing crack deep inversion model,which is the loess gully region mine The monitoring and governance of collapse cracks is of great significance.