中文摘要
雲是影響地球天氣系統的重要因素之一，但其型態變幻萬千，因此在對雲進行研究之前，須先針對雲參數進行探討，包括總雲量、雲速、雲層厚度及本文所要探討的雲高。
雲高的辨識，除了使用目測、探空氣球及光達觀測等傳統方法外，目前亦可應用衛星觀測影像配合遙測理論來進行雲分類之應用研究。由於衛星觀測資料可有效率地提供大範圍的訊息，故本文將使用新一代之MODIS( Moderate-resolution Imaging Spectroradiometer)衛星影像資料來進行雲高之分類。MODIS觀測資料的波長範圍從可見光、近紅外、中紅外到熱紅外線，共有36個頻道可供分析應用。
本文採用17個在雲辨識上比較常使用之頻道，並利用倒傳遞類神經網路(BackPropagation Networks；BPN)來進行分類。由於雲高之分類結果在驗證上具有較高困難度，因此本文將利用遙測理論所獲得雲頂溫度及雲高等結果，作為分類模式輸出目標值以及辨識結果驗證之參考。最後，本文以多筆資料來進行結果之測試比對，顯示此法具頗高之準確度及可行性。

Abstract
Cloud is one of the very important factors influencing the weather system of the earth. Since the cloud type in attitude fluctuates multifariously, discussions on the cloud parameter have to be conducted firstly in this thesis. These include “total cloud amount”, “cloud speed”, “cloud thickness”, and “cloud high”.
The classification of cloud high in traditional methods is performed by the estimating of human eyes, radiosunde, and liadar detection …etc. In addition, we can apply satellite images to collocate with remote sensing theory to implement the cloud classifies. Since satellites can offer the information on a large scale to observe the materials efficiently, a new generation''s MODIS (Moderate-resolution Imaging Spectroradiometer) is used in this thesis to obtain satellite image data to conduct the “cloud high” classification. The observation wavelength range of MODIST is from visible, near IR, middle IR to thermal IR, which has 36 channels suitable for analyzing and accomplishing the application.
In this thesis, 17 channels that are frequently used in cloud distinguishing are adopted, and the “Back Propagation Networks” (BPN) is employed for classification purpose. Since it is difficult to evaluate the results of cloud high classification, the cloud top temperature and cloud
high results generated from remote sensing theory are used to obtain the goal value of classification model outputs to distinguish the results for verification and consulting purposes. In our experiments, image data of different seasons were collected to test and compare the performance. Experimental results demonstrate that the proposed method is feasible and achieves very high accuracy rate in cloud classification

參考文獻
[1] I. S. Kathleen, E. G. Liam, R. Thomas, H. L. Huanga, and R. Dengel, “MODIS/AIRS instrument direct broadcast products and applications,” AMS 12th Conference on Satellite Meteorology and Oceanography, Feb. 2003.
[2] O. B. Brown and P. J. Minnett, “MODIS infrared sea surface temperature algorithm,” Tech. Report ATBD25, University of Miami, Miami, FL 33149-1098, 1999.
[3] M. P. Weinreb, R. Xie, J. H. Lienesch, and D. S. Crosby, “Destriping GOES M image by matching empirical distribution functions,” Remote Sensing of Environment, vol. 29, pp. 185-195, 1989.
[4] 呂維哲, 蘇木春, “模糊類神經網路為架構之遙測影像分類器設計”, 碩士論文,民國91年6月
[5] 林文賜, 周天穎, 林昭遠, “應用堅督性類神經網路於衛星影像分類技術之探討”, 航測及遙測學刊 第六卷 第一期 第41-58頁,民國90年7月.
[6] L. Bruzzone, D. F. Prieto, and S. B. Serpico, “A neural-statistical approach to multitemporal and multisource remote-sensing image,” IEEE Transactions of geoscience and Remote sensing, Vol. 37, No. 3, pp. 1350-1359, May 1999.
[7] J. A. T. Arriaza, F. G. Rojas, M. P. López., and M. Cantón, “An automatic cloud-masking system using backpro neural nets for AVHRR scenes,” IEEE Transactions of Geoscience and Remote sensing, Vol. 41, No. 4, pp. 826-831, April 2003.
[8] T. J. McIntire and J. J. Simpson, “Artic sea ice, cloud, water, and lead classification using neural networls and 1.6-μm data,” IEEE Transactions of Geoscience and Remote sensing, Vol. 40, No.9, pp. 1956-1972, Sep. 2002.
[9] B. C. Gao, P. Yang, and R. R. Li, “Detection of high clouds in polar regions during the daytime using the MODIS 1.375-μm channel,” IEEE Transactions of Geoscience and Remote sensing, Vol. 41, No.2, pp. 474-481, Feb. 2003.
[10] B. Tian, M. A. Skaikh, M. R. Azimi-Sadjadi, T. H. Vonder Haar, and D. L. Reinke, “A study of cloud classification with neural networks using spectral and textural features,” IEEE Transactions of Neural Networks, Vol. 10, NO. 1, pp. 138-149, Jan. 1999.
[11] H. Liu, V. Chandrasekar, and E. Gorgucci, “Detection of rain/no rain condition on the ground based on radar observations,” IEEE Transactions of Geoscience and Remote sensing, Vol. 39, NO. 3, pp. 696-699, Mar. 2001.
[12] 游丕若, “雲的奧秘-台灣篇”,人人出版社. 2002年5月.
[13] 游丕若, “雲的奧秘-圖鑑篇”,人人出版社, 2002年3月.
[14] 林文智, “雲與天氣”, 渡假出版社, 1996年6月.
[15] 林博雄, “賞雲”, 行政院農委會,民國89年11月.
[16] 陳正平, ”變色的天空”, 遠哲科學教育基金會, 1998年3月.
[17] 曾忠一, “大氣衛星遙測學”, 渤海堂文化事業公司, 民國77年.
[18] 曾忠一, “大氣遙測原理與應用”, 中央氣象局, 民國72年10月.
[19] 王寶貫, “雲物理學”, 渤海堂文化事業公司, 1995年7月
[20] 劉振榮, “遙測原理與應用”, 中央大學太空遙測研究中心氣象衛星實驗室.
[21] 國家太空中心, URL: Http://WWW.nspo.org.tw
[22] 中央大學太空及遙測研究中心URL: Http://www.csrsr.ncu.edu.tw
[23] 中大太遙中心氣象衛星實驗室.URL: Http://MSL.csrsr.ncu.edu.tw
[24] 蘇木春, 張孝德, “機器學習-類神經網路、模糊系統、以及基因演算法則,” 全華科技出版社, 民國91年3月
[25] A. Zaknich, “Neural networks for intelligent signal processing,” University of Western Australia, World Scientific Inc. 2003.
[26] 林昇甫, 洪成安, “神經網路入門與圖樣辨識,” 民國85年5月.
[27] 張斐章, 張麗秋, 黃浩倫, “類神經網路-理論與實務,” 東華書局,民國92年9月.
[28] 周鵬程, “類神經網路入門-活用matlab,” 全華科技圖書公司, 民國93年3月.
[29] 余建政,俞克維, 林義隆, 白能雄, “MATLAB 6.X使用入門,” 新文京開發有限公司,民國92年1月.
[30] 張智星, “MATLAB 程式設計與應用,” 清蔚科技出版社, 民國93年6月.
[31] S.Ackerman, K Strabala, P. Menze, R. Gery, B. Baum, C. Schaaf, and G. Rigger, “Discriminating clear-sky from cloud with MODIS algorithm theoretical basis document(MOD35),” MODIS Cloud Team, Ver. 3.2, Nov. 1997
[32] 葉怡成, “類神經網路模式應用與實作,” 儒林圖書有限公司, 2003年11月.
[33] 羅華強, “類神經網路-MATLAB的應用,” 清蔚科技出版社, 民國 90年.
[34] W. E. Shenk and R. J. Holub, “A multispectral cloud type identification method developed for nimbus 3 mrir measurements,”Mon. Weather Rev., vol. 104, pp. 281-291, Mar. 1976.
[35] M. Desbois and G. Szejwach, “Automated classification of clouds on meteosat imagery: Application to high-level clouds,”J. Clim. Appl. Meteorol., vol. 21, pp. 401-412, 1982.