隨著感測器整合以及衛星影像空間與輻射解析力之提升,以高解析度衛星立體影像對(Stereo-pair)產製空間圖資之潛力與日俱增。考量跳脫衛星影像公司於立體影像對供應之設限及提高立體影像對組成之選擇彈性,本研究嘗試以不同感測器之影像組成立體影像對,並建立物空間交會演算法則。 根據衛星影像物像對應供應方式,本研究工作發展兩種交會模式,一為以轉換後之嚴密感測器模式(Transformed Rigorous Sensor Model, TRSM)與有理函數模式(Rational Function Model, RFM)進行物空間交會解算;另一交會方法則為利用轉換後之有理函數模式(Transformed Rational Function Model, TRFM)與嚴密感測器模式(Rigorous Sensor Model, RSM)進行物空間交會解算。經理論分析及實驗證明,兩種方法在物像對應解算本質上是相同的,且適用於物空間三維定位解算任務。
Through the combination of sensors and the improvement of resolution of satellite imagery, the potential of using high-resolution satellite stereo-pairs to produce geo-spatial data has gained tremendous recognition. In order to get rid of the supplied restriction of stereo-pairs from vendors, and on the other hand, to improve the flexibility of establishing stereo-pairs, this research proposes a mathematical model for solving 3-D positioning allowing stereo-pairs that result from different sensors. Considering the types of orientation information provided by the satellite imaging vendors, the research developed two mathematical models for carrying on intersection computation. One of which is to integrate transformed rigorous sensor model (TRSM) with rational function model (RFM), and the other is to fuse transformed rational function model (TRFM) and rigorous sensor model (RSM). Based on the theoretical realization and the experimental analysis, it showed that the aforementioned models are identical and both suit performing 3-D positioning when dealing with inhomogeneous sensors.