無人機航空攝影測量輔助土地複丈可行性之研究

土地鑑界複丈關係人民福祉，依地籍圖測量及保存方式可分為數值區地籍圖、圖解區地籍圖、圖解數化區地籍圖三種。數值區地籍圖由於坐標系統已統一TWD97地籍坐標系，因此可直接利用數值法進行土地鑑界複丈；而圖解區地籍圖在近年來雖已逐漸數化，但數化後的地籍圖仍保存圖紙伸縮、褶皺破損之現象，這些現象會造成圖地不符、地籍圖分區接合、套圖前後面積改變等問題。因此鑑界前，需地測現況執行套圖分析。過去針對圖解數化區土地鑑界複丈進行之研究，大多利用地面測量方式對現況點進行測量，再將現況圖與地籍圖進行套圖分析。相較地面測量，使用無人機航空攝影測量能提供大量現況資訊，也可觀測地測難以觀測之現況點位。與現行圖解數化區土地鑑界複丈只針對欲鑑界土地附近進行局部套圖分析相比，利用無人機航測繪製之現況圖因含有大量現況資訊，可針對整體區域進行套圖分析，在整體區域內之界址點皆符合精度要求的前提下，減少為了提升欲鑑界單筆土地附近套圖精度而使整體精度下降之情形；且針對數值區土地鑑界複丈與圖解數化區土地鑑界複丈時，可利用無人機影像多光線交會之方式補建圖根點，增加自由測站法使用之彈性。因此，本研究於實驗區佈設合適之控制點與用eGNSS觀測其坐標後，利用四旋翼無人機酬載具55mm焦距中像幅量測型相機，於航高約300公尺取得前後左右80%且地面解析度3公分之影像，並以Pix4DMapper及Orima軟體進行空三平差，成果顯示在Pix4DMapper或將ORIMA空三解算成果輸出無畸變差影像於Pix4DMapper中進行圖根補點之可行性；而於套圖方面，以正射影像輔助辨識並以所測現況圖於ArcMap中執行Affine轉換對實驗區分區套圖，成果顯示套圖後地籍圖更符合土地現況，套圖前後現況點至地籍線垂距差異量減少2-9公分不等；最後再由實地測設方式完成複丈程序，測量成果均符合相關規範，顯示以無人機航測方式輔助土地鑑界複丈之可行性。

關鍵字

圖解數化地籍圖；土地鑑界複丈； UAV航空攝影測量

並列摘要

Survey of boundary verification affects people's well-being. According to the methods of surveying and the ways of preservation, there are three kind of maps, i.e. digital cadastral maps, graphic cadastral maps and digitalized graphic cadastral maps. The coordinate system of the digital cadastral maps have been unified into TWD97 cadastral coordinate system, so the surveyors can use digital method to perform boundary verification. Additionally, though graphic cadastral maps have been digitalized for the past few years, digitalized graphic cadastral maps still preserve differential shrinkage, wrinkles, folds and damage of the original maps. It will cause problems such as the difference between maps and the map of land detail situation, map join of the map sheets, and the change of the area after map registering. Therefore, map registration with land detail situation should be done before boundary verification. For most studies on the boundary verification of the digital graphic cadastral maps used ground survey to survey land detail situation to perform map registration. Compared to ground survey, UAV photogrammetry can provide lots of land detail situation. Moreover, the current map registration only locally focuses on the neighborhoods of the parcel for boundary verification. UAV photogrammetry can measure land detail situation globally to perform map registration in the whole area. Therefore, it can reduce the problem of local consistence from local land detail situations surveying by ground surveying and increase the reliability of the map registration globally. Additionally, for the boundary verification in digital cadastral areas or digitalized graphic areas, control points are probably lost. Using space intersection of aerial photogrammetry by multiple bundles to measure the coordinates of control points can bring more flexibility when using free station method for verification of parcel points. Therefore, this study used four-rotor UAV carried middle format metric camera with 55 mm focal length at flying height about 300 meters to collect the images with 80% endlap and sidelap and 3cm ground sampling distance. Before that, the suitable control points were targeted and surveyed by eGNSS. After that, Pix4dMapper and ORIMA software were employed to perform aerotriangulation, the result show that the control points for verification of parcel points can be measured in Pix4Dmapper by space intersection of aerial photogrammetry by multiple bundles or using the aerotriangulation result of ORIMA to generate undistorted images and import these images with related results into Pix4Dmapper for determining the coordinates of control points by space intersection of aerial photogrammetry by multiple bundles are possible. Moreover, the generated orthophoto can identify the land detail situation and the land detailed map of generated by UAV aerial photogrammetry can assist map registration using Affine transformation in ArcMap can make the cadastral map and land detail situation more consistent. The difference between land detail points to cadastral lines can be reduced by 2-9 centimeters after map registration. Finally, the parcel points can be stakeout to complete the process of boundary verification, and the result proves that UAV photogrammetry is of the feasibility to assist the boundary verification.