整合航測與行動裝置影像的三維建物模塑策略

Abstract

繼電子化政府、政府網路服務之後，標榜能察覺城市脈動、即時做出決策反應的智慧城市（Smart City），已成為世界各主要都市努力的目標。打造智慧城市必須先建置精確完善的三維建物模型，才能將各類感測器在三度空間中精確定位，而後才能對感測器傳回的資料進行空間分析，因此三維建物模型實為智慧城市之基礎建設。三維建物模塑主要可分為兩個階段：幾何模塑及牆面紋理貼圖。過去航空攝影測量一直是最有效且精確的測繪技術，然而傳統以浮測標逐點量測的方式重建三維建物模型時，操作員需要專注於屋角點的量測及點位之連結，是建物幾何模塑的瓶頸。為了使三維建物牆面具備真實的場景，必須花費大量人力、時間至現場拍照，回到內業還必須依據建物牆面逐一比對照片，挑出最完整的照片裁切出牆面影像，再依像片角度校正變形，才能作為牆面材質影像，需要大量人力、物力及時間。本研究以浮測模型（Floating Model）為核心，將傳統的浮測標擴充為浮測模型（Floating Model），使量測的單元不再是抽象的一個點位，而是許多種可伸縮、旋轉、移動的三維模型。浮測模型除了具備一個基準點可調整其三維位置外，還依模型種類增加了各方向的伸縮尺度、空間旋轉等參數，可想像成一個漂浮在空間中，可移動、旋轉、縮放大小以量測地物的模型。將浮測模型投影至所有影像，並套合至影像上的目標地物，便有如回復攝影瞬間之幾何空間條件，此時之模型參數與影像外方位元素即為最佳解。以此概念，本研究提出一套半自動化萃取策略，操作員透過人機互動介面將模型套合至所有航測影像，交由電腦透過最小二乘模型-影像套合演算法（Least-squares Model-image Fitting, LSMIF）自動計算最佳套合，得到最佳模型參數。其後將行動裝置拍攝建物的影像輸入電腦，由電腦自動計算每一面牆投影在所有像片上，以評估擷取牆面影像範圍，進而自動製作建物材質影像，完成三維建物模塑。透過實驗案例證明，浮測模型理論確可有效地萃取建物三維空間資訊，提高三維塑模之效率，並能達到傳統攝影測量之精度要求。

In the last decade, smart city, instead of e-government or i-government, has become a modern trend of urban development for the major cities in the world. However, a city would never be considered smart without knowing where the events are. A precise 3D city model is required as the infrastructure to mark and present the 3D location of the various sensors. The data transferred from the sensors can then be spatially analyzed in the 3D space. Photogrammetry has been considered as the most efficient technique for extracting 3D information or reconstructing 3D models. But its point-by-point measurement of using floating mark has become the bottleneck while reconstructing the 3D city model. In this paper, we expanded the floating mark to the floating model based on the concept of model-based building extraction. The measuring tool is no longer only an abstract point but also many kinds of 3D model, which can be scaled, rotated, or moved in the space. The floating model is defined with a datum point whose 3D coordinates indicate the spatial position of the model as the floating mark does. Furthermore, each kind of models is associated with a set of pose parameters to describe its rotation about the three orthogonal axes and shape parameters to describe its scales along predefined directions. In other words, the floating model is a flexible entity floating in the space, and can be adjusted to fit the object by these parameters. If the model parameters are good enough to represent the 3D spatial information of the object, the projection of the floating model on every overlapped image will all be coincident to the object's image. Therefore, the basic idea of the floating model theory is to fit model to the overlapped images by adjusting the model parameters. Based on the floating model theory, we proposed a semi-automated 3D building reconstruction strategy. A friendly human-machine interface is designed for the operator to choose and adjust the floating model to fit the aerial photos manually. Then, the computer calculate the optimal fit by an ad hoc Least-Squares Model-Image Fitting (LSMIF) algorithm. Thus the 3D spatial information can be extracted object-by-object rather than point-by-point by means of floating model, which increases the efficiency and accuracy. The building model are then projected onto photos taken by mobile devices to evaluate which is the most suitable photo for each fa蓷de. The fa蓷de then could be clipped from the photo and geometrically corrected as for the texture image. According to our tests and experiments, the proposed semi-automated strategy does increase the efficiency of 3D building model reconstruction.