The Application Research On Airborne LIDAR Data Processing And The 3D Material Stock Reconstruction

Accurate volume measurement of the material stock is not only related to the security and stability of manufacture,but also determines efficiency and overall competitiveness.The rapid,efficient and accurate stacking volume measurement plays an important role in the national economic development and scientific production of the enterprise.The inventory mostly is processed by hand or continuous weighing on the belt currently,which had a low degree of automation and only a low precision result.So it is difficult to adapt to the needs of large-scale measurement of stacking volume.This thesis combined the laser measurement technology with the automatic driving technology of UAV to design the volume measurement system and made a research on data processing method of the measurement point cloud.This thesis analyzed the composition and measurement principle of the airborne LiDAR system,and completed the communication between the laser scanning system,flight control system and data acquisition system.As the position of the aircraft acquired by the DGPS,the flight attitude measured by the IMU,and the ground information got by the laser system,the high-precision point cloud information could be obtained.On account of the large point number,the more noise,and the irregular distribution in point cloud data,it was hard for 3D stacking reconstruction and volume calculation.This thesis adopted the filtering algorithm based on the classification of noise to delete and smooth the noise point cloud.In order to accelerate the processing speed and reduce the influence on the volume calculation,This thesis simplified the point cloud by simplification algorithm,which the grid's size could change with the different density and sharp point cloud.Interpolation algorithm could be used to solve the problem caused by the inhomogeneous and missing fist-hand data.According to the triangular result of the surface of the pile,the volume calculation was completed by using the microelement method and achieved a good consequence.According to the research on the data processing above,the software system was completed by the Microsoft Visual Studio 2010,QT and PCL library.The accuracy of measurement point cloud was tested by the accuracy of the plane position and elevation through the flat experiment,and repeated error of the system was obtained through field test.The results showed the system was fast in measurement,short in measuring period,simple in operation,high in accuracy,and had a wide application prospect.