| 摘要: | 本論文介紹一種立體掃描系統，可以在短時間內掃描出物件立體形狀，再經由從建系統，將物件數據從組成原始立體物件。所提出的立體掃瞄系統，主要由一組或多組二極體線性雷射光和影像擷取器所構成。本研究是將線性雷射光採用質心法，將一物件作多點判讀後，將其從建成原始物件。我們廣泛的進行了一系列實驗，評估所設計的立體掃描系統的性能與掃描速度。 實驗設計成兩群組分別為誤差校正和解析度比較。誤差校正方面，由於觀測上距離不斷改變，圖形會以非線性關係發生形變，造成觀測上圖形失真，必須將其修正。解析度比較方面，不同解析度會造成掃描速度不同，但解析又會影響圖形精準度，必須有所取捨。解析度較低時所擷取點較少，在整體運作時間較短；當解析度較高時所擷取點上升，運作時間則較久。 在實驗中由於只使用單雷射，只能從單方向進行掃描，只可掃瞄物件特定一面情況，若用於球形物體上，會由於擷取資訊不足而造成物件失真。若改使用多雷射掃瞄，可於不同多方向同時進行掃描，可使失真情況大幅下降。 This paper describes a three-dimensional scanning system. It can scan the objects three-dimensional shape in a short time, and then through the build system, the object data can compose the original three-dimensional objects. This three-dimensional scanning system is composed by one or more sets linear diode laser light and video capture devices. The linear laser light is analysed by calculating the centroid in this experiment, the object will be a multi-point interpretation, then to built the original object. We conducted a series of experiments broadly, to assess performance and scanning speed of the three-dimensional scanning system. Experimental design into two groups: error correction and compared to resolution. Error correction, since distance will change during the observation, the image will deform by nonlinear relationship. It will distort on observation, so it need to be corrected. Compared to resolution, different resolutions will result in different scanning speed, and the resolution would affect the graphical accuracy too, so it need make a choice. Low-resolution, it will capture lesser points, so the time of operation is short; high-resolution, it will capture more points, it will need long time to operate. In the experiment, only from one direction to scan, so we only scan one side of the object. If it scans spherical objects, the object will distort because the retrieve information is lack. If it changes the use of multiple lasers, it can scan from different directions at one time, the distortions can significantly decrease. |
