本研究之目的為建置一套“無人機窄光譜遙測攝影系統”，以配合EcoNTU計畫之需求，在針對小範圍地區進行快速、準確且細緻的評估生態系統現況。然而一般光學遙感探測感測器需在晴空無雲的條件下，才能獲取高品質的影像資料。但這種條件在台灣這多雲的環境，一天中往往只有1至2小時的機會，因此如何有彈性的收集資料及有效的後處理程序乃是應用面需要考量的關鍵。本研究利用一架定翼型無人飛行載具(UAV)，裝置 MiniMCA-12窄波段多光譜相機以及單頻GPS天線協助定位。MiniMCA-12擁有12個鏡頭各自拍攝不同的波段，各個相機皆為框幅式相機，每台相機擁有自己的透視中心，因此若直接將所有波段進行套合將會造成波段錯位問題。MiniMCA-12所涵蓋的波段範圍為450~950奈米，每個波段的頻寬為10奈米，由於該相機所涵蓋的波段從藍光到近紅外光，與常用的生物物理指標很接近，因此適合用來對地表進行環境觀測。傳統上航空相片影像正射化過程中，首先需要進行連結點影像座標自動匹配，再搭配地面控制點進行(有或無)GPS/IMU輔助之空中三角測量。然而MiniMCA-12各個波段對地表反應之特性不同，無法在12個波段上匹配得到相同的地物點，此外若12個波段之影像一起進行空三平差，則影像張數為單一波段的12倍，影像總數及未知數太多，若使用傳統空三平差程序求解MiniMCA-12之外方位參數將會降低處理效率。
因此本研究提出兩種策略進行波段套合及影像正射化，策略一為假設MiniMCA之12台相機拍照時為同步攝影，先在室內實驗場針對12個鏡頭進行相機率定，同時求得12台相機各自之內方位參數，以及兩兩相機間的相對方位。接著在利用UAV進行航拍後，以其中一個相機當作參考相機，利用其拍攝之影像進行空三平差解算得參考相機之外方位參數。接著利用相對方位推求得其他相機的外方位參數，如此即可使用數值地形模型分別產生12個波段的正射影像。策略二為利用透視投影轉換，將各鏡頭像平面轉換至以鏡頭0為主之像平面，因此只需利用鏡頭0之內、外方位參數搭配DTM即可產生12個波段的正射影像。最後本研究以物件導向影像分析技術，針對MiniMCA影像進行初步地物分類研究，使用之特徵指標有波譜、紋理與形狀等，初步測試成果大致可分出陰影、建物、水體、植被、道路與裸露地等六大類。

The purpose of this research is to establish a “UAV narrow band multi-spectral remote sensing imaging system” to fit the requirement of EcoNTU project for rapid and accurate moniotring the conditions of ecosystems in great details over a small region. However, an optical remote sensor requires a clear sky condition. The window of opportunity is usually short (e.g. 1-2 hrs) in Taiwan, thus a flexible data collection and effective post-processing scheme is the key. In this study, a fixed-wing UAV equipped with a MiniMCA-12 narrow band multi-spectral camera and a single frequency GPS antenna is thus suggested. The MiniMCA-12 consists of 12 cameras and collect 12 channels of imagery individually. Each of them is a frame-based optical sensor with its own perspective center thus a direct overlay of all channels will induce band miss-registration problem. The bandwidth for each channel is 10 nm that covers from 400 nm (blue) to 950 nm (near infrared) spectral range with close regard to known biophysical indices suitable for environmental observation purposes. Conventionally, a precise ortho-rectification procedure requires tie-point image coordinates measurements, with/without GPS/IMU observations, and bundle adjustment with ground controls. However, those 12 bands of MiniMCA have different spectral responses for the same ground object, there is no chance to match the same object from all channels. In the meanwhile, if all 12 bands of images were used during bundle adjustment, the number of images will be 12 times to single channel only. The number of images and unknowns have increased dramatically that will cause in-efficiency by means of traditional aerial triangulation procedure. In this study, we thus propose two strategies for band-registration and ortho-rectification of MiniMCA camera. For the first one, we assume all 12 cameras of MinicMICA are synchronous. They were calibrated through an indoor calibration field including their interior orientation parameters (IOPs) and relative orientation parameters (ROPs) between any two cameras. After data collection through the UAV, we choose one channel as the reference image and perform aerial triangulation to obtain the EOPs of the reference one. Further, we can estimate the other 11 cameras’ EOPs based on the previous calibrated ROPs and combine a DTM for ortho-rectification of each band individually. In the second strategy, we apply the perspective transformation by reprojecting all cameras to the reference one on the image plane. Thus, all channels of images will have the same IOPs and EOPs as the reference camera and by means of DTM we can produce band-registered ortho-images for all 12 bands. In the end, we adopt object-based image analysis method for a preliminary study of MiniMCA image classification. The feature objects used includes spectral, texture, and shape indices. The experimental results show that shadow, building, water body, vegetation, road and bare land can be classified through the suggested classification scheme.

尹孝元、梁隆鑫、陳錕山、黃珮琦，衛星影像於國土變異監測之應用，航測及遙測學刊，第15卷，第1期，第 65-78頁，2010。
王鎮耀、吳建忠，利用衛星影像自動判釋應用於國有土地勘查之研究，2010。
吳究、張奇、劉銘哲，基於尺度不變特徵點轉換之精密多元影像套合，航測及遙測學刊，第14卷，第2期，第141-155頁，2009。
林士哲，空載三線式掃描儀影像立體定位之研究，國立中央大學，土木工程學系 ，碩士論文，2005。
林恩凱，利用高光譜影像偵測外來植物-以恆春地區銀合歡為例，國立中央大學， 土木工程學系，碩士論文，2005。
徐百輝，大地的辨識密碼-高光譜影像，科學發展 ，第416期，第13-19頁， 2007。
葉柏嘉、饒見有，以線段偵測為基礎之半自動三維房屋模型重建方法，國立成功大學，測量及空間資訊學系，碩士論文，2012。
魏子軒，高光譜影像分類辨識方法，儀科中心簡訊，2007。
羅秋月，IKONOS衛星影像正射改正之研究，國立中央大學，土木工程學系，碩士論文，2002。
Bay, H., Tuytelaars, T., Gool, L.V., Zurich, E., 2008. Speed UP Robust Features (SURF). Computer Vision and Image Understanding 110, 346-359.
Broge, N.H., Leblanc, E., 2000. Comparing prediction power and stability of broadband and hyperspectral vegetaion indices for estimation of green leaf area index and canopy chlorophyll density. Remote Sensing of Environment 76, 156-172.
Chen, Y.H., Su, W., Li, J., Sun, Z., 2009. Hierarchical object oriented classification using very high resolution imagery and LIDAR data over urban areas. Advances in Space Research 43, 1101-1110.
Jhan, J.P., Rau, J.Y., Lin, C.C., Shao, Y.C., 2011. Automatic Tie-point Matching for Semi-automatic Photo-triangulation. 第三十屆測量及空間資訊研討會, 9月1-2日,國立中興大學, 台中市, 台灣.
Kelcey, J., Lucieer, A., 2012. Sensor Correction of a 6-Band Multispectral Imaging Sensor for UAV Remote Sensing. Remote Sensing 4, 1462-1493.
Laliberte, A.S., Goforth, M.A., Steele, C.M., Rango, A., 2011. Multispectral Remote Sensing from Unmanned Aircraft: Image Processing Workflows and Applications for Rangeland Environments. Remote Sensing 3, 2529-2551.
Lillesand, T.M., Kiefer, R.W., 2000. Remote Sensing and Image Interpretation. John Wiley & Sons, Inc.
Lowe, D.G., 2004. Distinctive Image Features from Scale-Invariant Keypoints. International Journal of Computer Vision 60, 91-110.
Mikolajczyk, K., Schmid, C., 2005. A Performance Evaluation of Local Descriptors.
Pu, R., Landry, S., 2012. A Comparative Analysis of High Spatial Resolution IKONOS and Worldview-2 Imagery for Mapping Urban Tree Species. Remote Sensing of Environment 124, 516-533.
Tang, L., Dörstel, C., Jacobsen, K., Heipke, C., Hinz, A., 2000. Geometric accuracy potential of the Digital Modular Camera. IAPRS ⅩⅩⅩⅢ.
Torres-Sanchez, J., Lopez-Granados, F., De Castro, A.I., Pena-Barragan, J.M., 2013. Configuration and Specifications of an Unmanned Aerial Vehicle (UAV) for Early Site Specific Weed Management. PLoS One 8, e58210.
Wolf, P.R., Dewitt, B.A., 2000. Elements of Photogrammetry： With Applications in GIS. McGraw-Hill, Boston.
Yeh, P.C., Rau, J.Y., 2012. A Semi-Automatic Image-Based Close Range 3D Modeling Pipeline Using a Multi-Camera Configuration. Sensors (Basel) 12, 11271-11293.