本研究使用逆向工程的方法搭配MicroScribe G2進行真人頭顱量測建立電腦模型,其作法爲使用矽膠彈性土建立頭顱負模後使用MicroScribe G2取點建立電腦模型,之後產出RP(rapidprototype)的安全頭盔,並請受測者穿戴。爲了解哪些因子影響電腦模型和實際物體的密合度,以相親相愛木雕作樣本並以田口L16實驗作實驗和分析,實驗的4 個因子爲:操作人員、SVD(singular value decomposition)的點數、結合點重覆量測的次數和檢驗點的重複次數。量測的標準爲檢驗點差異的SN比。SN比越高表示密合度愈好。本實驗指出操作人員、SVD 的點數和檢驗點的重複次數爲顯著因子。其中SVD的點數爲6時和檢驗點的重複次數爲5時產生的密合度最佳。我們用SN比所產生的S/N ratio得知前述兩個因子的水準也能產生最穩定的結果。
Reverse engineering with MicroScribe G2 was used to measure the negative mold of a human skull to construct a computer model thereof in Rhino, a task accomplished by using special clay to obtain a negative impression of the skull and measurement by MicroScribe G2 to build the model. Later a rapid prototype (RP) of ABS (acrylonitrile butacliene styrene) material was created for a helmet to fit the head studied. To understand the factors influencing the match between the computer model and its artifact, ”Love each other,” a wooden sculpture, was used as a sample to analyze data obtained by Graceo-Latin Square Design and the Taguchi method. The experiment showed that the significant factors were the operators, the SVD (singular value decomposition) number and repetitions of check points.