火星HiRISE數值地形模型之產製及應用於火星地表探測車定位之研究

Abstract

火星地形資料為火星探勘工作的基礎，一旦能夠掌握火星的地表情形，便可透過地形判讀，推論火星過去的地表活動、氣候變遷與太空生物演化等課題。因此，本計畫首要目的即針對高解析度HiRISE 火星影像，建置台灣第一套HiRISE 影像處理系統，製作正確且高解析度之數值地形模型（DTM）以及正射影像，以提供之後國內外專家學者在火星相關課題之研究使用。本計畫的另一個研究重點為多重衛載DTM 之間，以及衛載DTM 與地面移動站（MER）資料的整合，其中衛載DTM 的部分（包括MOLA、HRSC 與HiRISE DTM），將採用三維表面套合（Surface matching）方法，完成此三種DTM 的地理對位。套合成果可讓使用者在同一基準上，同時觀測同一地區多重解析度之地形資料，將有助於火星地形分析。下一階段將利用已完成對位之HiRISE DTM，配合MER 所拍攝的影像，採用自率光束平差法（Self-calibrating bundle adjustment）率定MER 像機，並重新定位 MER 像機在拍攝時位於火星表面的位置。最後，本計畫將編寫一套整合MER 以及HiRISE DTM 之系統，以規劃MER 前進至下一目標站的最適移動路線，減低移動過程中因遭遇地形起伏，導致移動站或是搭載儀器毀損之機率。綜上所述，本計畫之研究目的包括：（1）建置火星HiRISE 影像與DTM 處理系統。（2）率定MER 像機並定位MER 於火星表面位置。（3）MER 最適前進路線之規劃系統。

Martian topographic data is an essential element for the scientific exploration of Mars, such as geological analysis, geomorphological interpretation, climatic and potentially astrobiological evolution, and landing site selection, etc. To fulfil increasing demands from geoscientists worldwide, a number of imaging systems (such as High Resolution Stereo Camera (HRSC) and High Resolution Imaging Science Experiment (HiRISE)) onboard various Mars orbiters have been collecting imagery of Martian surface over the last decade. Due to the high resolution of HiRISE image (25 cm per pixel), this research mainly focus on establishing a HiRISE image processing system for generating high resolution digital terrain models (DTM) and ortho-image. The results are expected to meet the demands for scientists who are interested in Mars research. In addition to the HiRISE data, there are a number of multi-sensor, multi-resolution and multi-scale Martian topographic products available. In order to fully utilise these Martian datasets, other important topics of this research is to co-register Mars DTMs produced from multiple orbiter data (including Mars Orbiter Laser Altimeter (MOLA), HRSC and HiRISE DTMs), and also an integration between orbiter DTMs and ground image data obtained by Mars Exploration Rovers (MER). Surface matching is the core technique to implement the former task. Once the matching is finished, the HRSC and HiRISE DTMs are co-registered to the MOLA DTM. As they have been co-aligned based on the same reference, the datasets can be observed or exploited simultaneously. At next stage, a self-calibrating bundle adjustment is applied to calibrate MER camera. For the calibration, the images taken by MER camera are used and the co-registered HiRISE DTM is involved as the geodetic control information. Once the MER camera is calibrated, the location of the rover is also re-derived. In the end, a system combining MER and HiRISE data will be developed, in which the location of the MER can be illustrated. Additionally, together with the topographic information extracted from the HiRISE DTM (such as slope, surface roughness, etc.), potential routes for MER to traverse toward next station can be plotted. In summary, the objectives of this research are: （1）Establish a Mars HiRISE image and DTM processing system. （2）Calibrate MER camera and re-solve the MER location on Mars terrain surface. （3）Develop a system capable of illustrating MER location and potential traverse route