| Soil erosion is a global environmental problem,which our government attaches great importance to.We stipulate that when carrying out development and construction,it is necessary to predict the possible soil erosion.At present,soil erosion monitoring methods mainly include land filling method,3D laser scanning method and topographic pinplate method.Due to the lack of some basic data and theoretical research,it is difficult for these methods to meet the requirements of quantitative,scientific and accurate.Therefore,it is one of the important problems that need to be solved urgently to find new soil erosion monitoring methods.Photogrammetry technology is characteristiced by non-contact measurement method,no damage to the measured subject,abundant image information,easy storage of informnation,reusable information,high measurement accuracy,low cost,fast speed,low outdoor labor intensity and so on.In recent years,it has been applied in geology,water conservancy,transportation and other fields.Therefore,the application of photogrammetry to soil erosion measurement has the far-reaching significance.To study the precision and influence factors of oblique photogrammetry in monitoring soil erosion.In this study,the differences in erosion amount between conventional methods and oblique photogrammetry were studied through the laboratory erosion test,and the calculation accuracy of different oblique photogrammetry techniques was explored,and the accuracy of oblique photogrammetry was analyzed in field in-situ slope test.A comparative analysis of DSMs generated by ground image collection of SLR camera and low-altitude image collection of unmanned aerial vehicle was made to explore the impact of different resolutions on measurement accuracy.The result of research showed:1.The average error of Pix4D was 13.43%and the average error of Smart 3D was 15.69%,so Pix4D was better and had higher accuracy.Besides,by setting the minimum threshold,the error of DSM generated by SFM-MVS,close-range photogrammetry on soil with minimal erosion could be well corrected.The total mean error of DOD(Digital surface Model of Differences)was 10.00%after the use of threshold in undisturbed colluvial soil,which was 7.16%less than that of the unused threshold.Therefore,the minimum threshold setting improved the measurement accuracy of DOD and made the measurement results more stable.2.The image processing results of the unmanned aerial vehicle showed a good matching effect of SFM low-altitude oblique photogrammetry technology,with the total mean re-projection error was about 0.47 um and the total root-mean-square positioning error of ground control points was only 3mm.Through the analysis of the transport process of soil erosion and sediment in the study area in the two periods,the main erosion and sediment changes in the rill in the two standard runoff zones were obtained,and the changes of soil erosion and sediment transport process could be obviously obtained under the condition of small rainfall.The experimental results showed that low-altitude oblique photogrammetry was feasible for monitoring outdoor standard runoff plot.3.The mean re-projection error of photogrammetric measurement at different intervals during the same period was about 0.2 pixels,and the error of ground control point was close to the result of man-machine low-altitude oblique photogrammetry.After the point cloud data was superimposed,it was found that,with the increase of the shooting distance,the superimposed effect of resultant point cloud and the point cloud generated by unmanned aerial vehicle was reduced,and the average error of Gauss model was increasing,and the maximum mean elevation error was up to 8.003mm,with point cloud reproducibility poor.The photogrammetric method of image acquisition around the standard runoff plot at different shooting intervals had a poor effect on the monitoring process of soil erosion and sediment transport,therefore,it is not recommended to use this method in the monitoring of outdoor standard runoff plot. |
PDF Full Text Request