Research On Application Of Dangerous Rock Monitoring Based On UAV Image Matching Point Cloud Data

According to the national "14th Five Year Plan" and Tibet’s medium to long-term railway plan,Tibet will build and renovate more than ten railway trunk lines.The large-scale construction and development of railways can lead to chain geological disasters.Rock mass undergoes deformation under geological processes and external forces,posing a threat to roads,railways,and surrounding personnel.The problem of a large number of bridge tunnel ratios and the inability of personnel to reach is prominent.Drones have the advantages of convenience,speed,and wide coverage,which can replace personnel in dangerous rock monitoring.In this context,this article selects the unavoidable dangerous rock disasters in the construction and reconstruction of railway main lines as the object,and mainly studies the technical application method of using drone image matching point cloud data for dangerous rock monitoring.The research is conducted through campus experiments,field experiments,and engineering practice.Using campus experiments,summarize the methods and techniques for generating point cloud models using software such as drone butler,Context Capture Center Master,and Inertial Explorer.Compare the functions,techniques,generation methods,and accuracy of drone image matching point cloud models and airborne radar generated point cloud models.In the experiment,RTK control point data was used as the reference point for accuracy analysis,and it was found that the average mean square error of the drone matching point cloud model was smaller than that of the airborne radar point cloud model.Therefore,a technical system for monitoring dangerous rock movement based on drone 3D modeling and point cloud recognition was established.Verify the application of drone image matching in dangerous rock monitoring through field experiments.The experimental data is visualized using Trimble Real Works and cloud comparison methods for point cloud comparison to monitor the displacement changes of dangerous rocks.The test results can clearly measure the distance and range of dangerous rock changes.Implement gradient filtering algorithm,CSF filtering algorithm,and Bilateral filtering algorithm through cloud compare software,and compare the imaging results of the filtering algorithms.By participating in the disaster monitoring work of the second transverse tunnel of the Layue Tunnel near the Qianlonggou area,the research content will focus on the method of generating point clouds for drone image matching;Using the point cloud data generated by the UAV tilt image,the dangerous rock displacement change monitoring method and other means are used to regularly monitor the dangerous rock Deformation monitoring at the upper end of the tunnel,and other sensor monitoring methods are used to monitor the dangerous rock and disasters during the actual construction of the Sichuan Tibet Railway.This study explores a feasible application method for monitoring dangerous rocks by studying drone 3D modeling and point cloud data.It uses point cloud data generated from drone tilt images to determine the distance of dangerous rock movement in areas that are difficult to reach and observe.By studying the reduction of economic losses and threats caused by dangerous rocks to railway construction and future operations,it also provides a new approach for the application of drone point cloud data.