The purpose of this study is to suggest the method of building registration in the current cadastral re-survey project and to derive limitations. And based on the derived limitations, we proposed a building registration method using UAV 3D modeling.
...
The purpose of this study is to suggest the method of building registration in the current cadastral re-survey project and to derive limitations. And based on the derived limitations, we proposed a building registration method using UAV 3D modeling.
In order to achieve the purpose of this study, the basic theories and general contents of building registration and UAV 3D modeling in the cadastral re-survey project were reviewed. In addition, this study attempted to grasp the current status of registration, management and utilization of existing building registrations and to derive limitations. In order to overcome the limitations, the building registration in the cadastral re-survey project was tested using the UAV 3D modeling technique. And the improvement plan was suggested by reflecting the experimental results.
The study area for the experiment was selected as a rural housing area in Nowon-ri, Jincheon-gun, Chungcheongbuk-do. This area is a mixture of buildings with and without building roof edges, and difficult to access buildings.
3D modeling image data of the study area was acquired by UAV surveying, and data processing was done using Context Capture S/W. In order to analyze the accuracy of the 3D modeling data, the RMSE of ground control point and verification point were calculated. As a result, the ground control point showed RMSE of X = 0.015m, Y = 0.012m, and connection intersection = 0.026m. The verification point was X = 0.015m, Y = 0.024m, and connection intersection = 0.028m, which means that the image can be used.
A total of 64 boundary points of the outer wall of the building in the experimental area were extracted by using the produced images and compared with the existing building status map and ground survey (TS) results. For accurate analysis, the classification of the visually discernible point and the limitation of the clear extraction were classified separately without being affected by the shade occlusion area and the structure of the building. As a result of classification, the total number of applicable points was 56 points, and the RMSE of these points was X = 0.035m, Y = 0.036m, and connection intersection = 0.051m. This may be applicable because it satisfies the scope of performance recognition of the boundary point prescribed in the current Special Act on Cadastral Re-survey.
However, as a result of comparing and analyzing the building status map and accuracy for 8 points, which were limited in boundary point extraction, RMSE was analyzed as X = 0.869m, Y = 0.590m, and connection intersection = 1.05m. For these buildings, building boundaries should be extracted in connection with ground surveying.
In this study, the legal, institutional, and technical improvement measures were suggested based on the experimental results for the registration of buildings using UAV 3D modeling techniques. This was divided into the enactment of legal regulations, the diversification of registration methods, and the establishment of a basis for maximization of utilization. It is expected that more efficient building registration will be possible if the method proposed in this study is applied in connection with the current ground surveying method.