UAV LiDAR 기반 Point Cloud 데이터를 이용한 토지경계 설정 연구

UAV LiDAR 점군 데이터를 이용한 토지경계 설정을 연구했다.

In cadastral surveying, land boundary extraction and establishment is divided into electronic plane table, GNSS, and UAV methods. Currently, the electronic plane table survey method is used for most land boundary establishment, but in the case of forest areas, it is not efficient in terms of manpower and time because it needs to be measured by moving relatively many station mark and short distances due to terrain characteristics such as large elevation differences and the influence of trees.

In the case of GNSS survey method, it is still not widely used in the field of cadastral surveying due to the multi-path phenomenon caused by satellite signal disturbance in urban areas where obstacles exist, and satellite signal blocking due to trees in forest areas. The UAV method suffers from various obstacles such as noise, shaky focus, obstacles and shading, depending on weather conditions, causing time and economic losses. For these reasons, there is a need for a land boundary detection method that incorporates other new technologies.

In this study, we aimed to acquire the location information of land status boundary points using UAV LiDAR-based point cloud data, which integrate the advantages of UAV, which can acquire rapid spatial information even in inaccessible areas, and LiDAR, which can scan precise terrain even in forest areas and reduce shadowing due to weather and shadows as much as possible, for efficient land boundary setting work. To achieve the purpose of this study, the basic parts of UAV and LiDAR were researched in advance, and the basic theories and general contents of UAV LiDAR and land boundaries that integrate these two technologies were reviewed.

In addition, we identified the laws on land boundary setting, the current status of land boundary setting by existing methods, and the application cases of UAV LiDAR, and tried to derive the limitations of these. To overcome the limitations, we designed an experiment and selected an area for the actual experiment. Then, we conducted an experiment to extract and set land boundaries using UAV LiDAR-based point cloud data. The experimental area was selected in Iwol-myeon, Jincheon-gun, Chungcheongbuk-do, where agricultural, residential, and forest areas are appropriately distributed and the boundaries between land and forest are clearly visible.

For the point cloud image used in the experiment, first, an orthophoto image was created using a fixed-wing UAV, and then a point cloud image was created using UAV LiDAR. For ease of analysis, RGB data was entered into the point cloud in conjunction with the orthophoto image. Then, land boundary extraction was performed using the method prescribed by the current law, and 1,014points were extracted, and 167points that could be verified by overlapping with the existing status map were separately extracted, and the intersection of the existing status map and the connection was compared and verified using the RMSE technique.

The results showed X=0.0758m, Y=0.1270m, and connecting intersection=0.1479m, and excluding the overestimation point, the results were X=0.0460m, Y=0.0757m, and connecting intersection=0.0886m. This is equivalent to the accuracy of ±0.07m for the cadastral re-survey and ±0.10m for the digital area. In the case of forest areas, since there was no comparative verification target, direct verification points were obtained by GNSS surveying for ridges, valleys, and the boundary between land and forest, and analysed the number of point clouds located within a 7cm radius around the point.

As a result of the analysis, an average of 46 points were located, proving the feasibility of its use in forest areas. Based on the above experimental results, we proposed an overall improvement plan for the current land boundary extraction and setting process. In order to apply the experimental results of this study to practice, first, legal amendments were required. In order to establish a legal basis for the application of UAV LiDAR, some laws that stipulate the method of cadastral surveying and the method of land boundary extraction were proposed to be amended to allow the use of UAV LiDAR.

In addition, amendments to the ⌜UAV cadastral survey regulations⌟ for the use of UAV LiDAR in cadastral survey and cadastral re-survey projects and the preparation of UAV LiDAR work process guidelines were suggested. Secondly, it proposed a plan to overcome the limitations of existing and new technologies by linking the existing land boundary extraction and establishment methods, electronic flat panel survey method and satellite survey method, and presented the necessity of updating the cadastral status map based on this.

Third, in order to improve the land boundary extraction method using UAV LiDAR, we proposed a plan to improve the extraction limit area through the integrated use of UAV and LiDAR, a technical improvement plan for the decrease in accuracy due to the shooting altitude, which is the biggest disadvantage of UAV, and a plan to introduce automation technology in conjunction with artificial intelligence technology for consistent land boundary extraction, suggesting the possibility of future development in the cadastral field. In conclusion, by applying UAV LiDAR to the process of land boundary extraction and establishment, we were able to achieve good results beyond our expectations.

In the case of the experimental part of this study, it is judged that it can be applied in practice by linking the current ground survey and cadastral map. UAV LiDAR technology is expected to have many advantages that can be used in the future to create stereoscopic cadastral maps and terrain modelling through 3D cadastral survey. 지적측량에서 토지경계 추출 및 설정은 전자평판, GNSS, UAV 방식으로 구분되어 활용되고 있다. 현재 전자평판측량 방식이 대부분의 토지경계 설정에 이용되고 있지만, 임야지역의 경우 많은 고도 차와 수목의 영향 등 지형적 특성으로 비교적 많은 기계점과 짧은 거리를 이동하며 측정하여야 하기 때문에 인력 및 시간적인 측면에서 효율적이지 못한 것으로 나타났다. GNSS 측량 방식의 경우에는 장애물이 존재하는 도심지에서 위성 신호의 교란에 따른 멀티패스(MultiPath) 현상과 임야지역에서는 수목으로 인한 위성신호 차단