Extraction Of Linear Structure Of Gaosong Ore Field In Gejiu Mining Area And Comparative Study Of Its Spatial Variabilit

The application of remote sensing technology in geological research is mainly reflected in the extraction of linear structures,which are mostly linear image features of fault structures on remote sensing images.As an important research object in geology,the distribution law of fault structure is an important symbol to identify deposits,mineralized zones and ore bodies.By extracting the linear structure of remote sensing images,the spatial distribution of fault structures can be effectively revealed.At present,most of the traditional linear structure extraction methods are based on the edge detection technology of digital filtering to extract the mitochondria in remote sensing images.These algorithms generally cannot extract the mitochondrial features in the shadow area of the mountain and the parallel area of the solar incident angle.In this paper,the STA algorithm(line segment tracking method)is selected to extract the linear structure of the remote sensing image in Gaosong ore field of Gejiu mining area,and the shortcomings and shortcomings of the linear structure are extracted according to the line segment tracking method(the turning angle is too large or the interval distance is slightly larger,the linear structure extraction result will appear the phenomenon of disconnection;the STA algorithm is improved if the line element interval of the same fault structure is too far.The improved algorithm is used to extract the linear structure.The main understanding and achievements are as follows :(1)On the basis of image preprocessing,geometric correction and image enhancement,the traditional Canny algorithm is used to extract the linear structure of the remote sensing image in the mining area.(2)The line segment tracking method is improved,and a fixed pixel point in the remote sensing image is innovatively proposed as the line element search center point.The 11 × 11 line element search window is set,and 16 equal angle and equal interval radial search directions are set in the range of 360° around the center point.The line element is determined within the search range,and then the line elements determined as linear structures in the direction with good pixel value continuity are connected according to a certain standard we set.The line element is used as the center point,and the line element connection direction is set(32 directions of data search are used as boundaries).Taking the direction with good continuity of radius and pixel value as the central direction,the image is scanned in the top corner fan-shaped area,and the midpoint line element is connected with the farthest line element in the area to obtain the linear structure extraction result.According to the results,it can be seen that the linear structure extraction results are more continuous,and due to the wider range of line element determination,the extraction results are more abundant,which can show the fracture structure of the mining area more comprehensively.(3)Two research areas in Gaosong ore field of Gejiu mining area are selected,which are located near Lianhuashan fault and Wuzishan anticline respectively.UAV images are taken and appropriate remote sensing images are selected in the research area.The traditional Canny algorithm is used to extract the linear structure in the image,and the fracture structure in the digitally extracted UAV image is used as the experimental control group,which is compared with the linear structure extraction results based on the improved line segment tracking method.The comparison method is : using the improved circular scanning line method,Arc GIS and geostatistical software were used to calculate the fracture strength and fracture density of the fault structures extracted by the traditional Canny algorithm,the improved line segment tracking method and the digital extraction of UAV images.Through comparative study,it is found that the attribute values(mean,variance,standard deviation and coefficient of variation)of the fracture strength and fracture density of the linear structure extracted by the improved line segment tracking method are closer to the digital attribute data of the fracture structure of the UAV image compared with the fracture attribute values of the linear structure results extracted by the traditional Canny algorithm.This shows that the linear structure results extracted by the improved line segment tracking method can more truly reflect the distribution of fracture structures in the mining area.(4)Compared with the traditional Canny algorithm,the improved line segment tracking method and the digital extraction of UAV images,the three sets of data of fracture structure show the spatial variability of fracture structure,and carry out field verification.The results show that the spatial variability of the measured data and the semi-variance function model established by the three groups of remote sensing image fault structure attributes show that the closer the fault structure is to the Lianhuashan fault,the more perpendicular the dominant joint development direction is to the Lianhuashan fault strike,and the maximum variability direction of the fault structure is almost parallel to the Lianhuashan strike.The fault structure is close to the Wuzishan anticline area,and the dominant joint development direction is more parallel to the Wuzishan anticline axis.The maximum variability direction of the fault structure is almost perpendicular to the Wuzishan anticline axis.Comparing the variability of fracture structure in three groups of images,it is found that compared with the traditional Canny algorithm,the spatial variability of fracture structure shown by the improved line segment tracking method is closer to the UAV measurement data and field measurement data.