Research On Method Of Sand Volume Measurement Based On UAV Aerial Photogrammetry Technology

In the construction of water conservancy,port and offshore projects,a large amount of sand is needed as backfilling and building materials,while sand is mainly from inland and needs to be transported from the sand yard to the construction site by sandship.Therefore,the sand volume measurement has always been the focus of the project construction,and it's efficiency and accuracy will directly affect the construction progress and cost control of the project.However,due to factors such as hull structure and water environment,commonly used earthwork methods on land,such as total station method and RTK method,are difficult to be applied to the sand volume measurement.At present,the common method for the sand volume measurement is to simplify the shape of the sand contained in the sandship to a prism or a cone,and then measure it with a tape and a level.This method requires multiple people to work together,and is susceptible to human factors,and the measurement accuracy is not high.In order to improve the efficiency and accuracy of the sand volume measurement,this thesis studies the method of the sand volume measurement based on the UAV aerial photogrammetry technology.The main contents include the following aspects:(1)Summarizes the UAV aerial photogrammetry technology,including system composition,data acquisition process and data processing method,and analyzes the advantages of the UAV aerial photogrammetry technology applied to the sand volume measurement.(2)Studies three key problems in the calculation of the sand volume based on UAV images.Aiming at the problem of distortion in the image acquired by the UAV,the camera is checked by Zhang Zhengyou's method in order to obtain the calibration parameters.The image distortion is corrected by indirect method.Aiming at the problem of image texture duplication of the sandship,the phase matching-assisted image matching method is used,and the good result is obtained.The accuracy of the final generated digital surface model(DSM)is checked,and the results show that the accuracy of DSM data meets the requirement of the sand volume measurement.Aiming at the problem that the DSM data generated by the UAV image is large and scattered,the ordinary Kriging interpolation method is used to interpolate the DSM into regular grid data,and then the final square is obtained by elevation difference between two sets of regular grid data.The results are compared with those calculated by Southern CASS software.The results show that the relative deviation between the two methods is less than 1%,which indicates that the calculation method based on the UAV image is reliable.(3)Two experiments were carried out to validate the method of the sand volume measurement based on the UAV aerial photogrammetry technology.The first experiment was carried out on land,using the UAV aerial photogrammetry technology to measure a sand pile,and compare it with the results of three-dimensional laser scanning.The relative deviation between the two is 0.85%,indicating that the UAV aerial photogrammetry technology is feasible for the sand volume measurement.The second experiment was carried out on the sandship.The five sandship were measured twice using the UAV aerial photogrammetry technology.The results of the two UAV aerial surveys were compared and found that the relative deviation was within 0.5%,which indicated that the UAV aerial photogrammetry technology had good stability in the sand volume measurement.Meanwhile,the average of the two UAV aerial surveys results was compared with the results of three-dimensional laser scanning.It is found that the relative deviation is less than 1.5%.This shows that UAV aerial photogrammetry technology can be used to measure the sand volume with the same accuracy as three-dimensional laser scanning measurement,which can effectively guarantee the measurement accuracy.(4)The influence of the average spacing of DSM point clouds and the grid spacing set during interpolation on the volume calculation accuracy is analyzed by using the methods of the one-way ANOVA and linear regression analysis.The results show that in order to ensure the accuracy of square calculation,the average distance between point clouds should not be more than 0.5 m and the distance between grids should not be more than 0.6 m.