臺灣地區易受到颱風暴雨侵襲，常發生水災，且因地形山高河短，淹水過程相當短暫，又受天氣影響，不適合使用飛機或由平流層上方的可見光衛星觀測完整災區，故最適合調查災區淹水區域之手段即為使用可穿透雲層之合成孔徑雷達（SAR）衛星。
本研究使用倒傳遞類神經網路（BPNN）模式及多變量線性迴歸（MLR）模式，結合SAR衛星影像資料，以建構淹水區域辨識模式。其中，BPNN模式可分為僅用淹水時SAR影像之模式一，以及使用淹水前及淹水時兩張SAR影像之模式二，輸入變數包含各像素點經轉換後之雷達散射截面（RCS）值、自身及其鄰近9宮格之統計平均值、標準差、最小值及最大值；MLR模式使用淹水前及淹水時兩張SAR影像，輸入變數兩張影像中各像素點之RCS值差異量、自身及其鄰近9宮格RCS值差異之統計變異數。
結果顯示BPNN模式有較佳的辨識效果，訓練資料與測試資料之淹水辨識正確率分別高達80%與73%以上。錯誤辨識區域大多為分佈零散、未集中於特定區域；為修正這些小而零散的區域，使用型態影像學運算處理將模式輸出結果進行修正，修正後結果正確率大為提升，辨識正確率可提升至90%以上。

Typhoons and storms hit Taiwan several times each year and they cause serious
flood disasters. The rivers are short and steep, and their flows are relatively quick
with floods lasting only few hours. Due to the factors of the weather, it is not suitable
for aircraft or traditional multispectral satellite; hence, the most appropriate way for
investigating flood extent is to use Synthetic Aperture Radar (SAR) satellite.
In this study, back-propagation neural network (BPNN) model and multivariate
linear regression (MLR) model are constructed to identify the flood extent from SAR
satellite images. The input variables of the BPNN model are the pixel’s Radar Cross
Section (RCS) value and mean, standard deviation, minimum and maximum of RCS
values among its adjacent 3×3 pixels. The MLR model uses two images, including
the flooding before and the input variables of the MLR model are the difference
between the RCS values of two images and the variances among its adjacent 3×3
pixels.
The results show that the BPNN model can perform much better than the MLR
model. The correct percentages are more than 80% and 73% in training and testing
data, respectively. However, the locations of many misidentified areas are very
fragmented and unrelated. For correcting the small and fragmented areas,
morphological operations are used to modify the outputs of these three identification
models. The modified results have been improved a lot and the correct percentages
increase up to 90%.

目錄
目錄.................................................I
表目錄...............................................III
圖目錄...............................................IV
一、前言.............................................1
1.1 研究動機與目的...................................1
1.2 研究方法.........................................2
二、文獻回顧.........................................3
2.1 應用類神經網路進行影像辨識.......................3
2.2應用類神經網路進行遙測影像分類....................3
2.3應用合成孔徑雷達衛星辨識水體或淹水區..............4
三、理論概述.........................................6
3.1 類神經網路.......................................6
3.1.1 倒傳遞類神經網路...............................8
3.1.2 誤差倒傳遞演算法...............................9
3.2衛星遙測概述......................................12
3.2.1合成孔徑雷達衛星................................13
3.2.2合成孔徑雷達衛星影像............................14
四、研究案例.........................................16
4.1 研究區域概況.....................................16
4.2 資料蒐集與處理...................................18
4.2.1 衛星影像處理...................................18
4.2.2 淹水區域調查...................................22
4.3評估指標..........................................26
4.4 模式架構.........................................28
4.5 淹水辨識模式.....................................33
4.6 綜合討論.........................................42
4.7以型態影像修正模式辨識結果........................45
4.8 宜蘭縣其他地區淹水區域辨識.......................48
五、結論與建議.......................................51
5.1 結論.............................................51
5.2建議..............................................52
六、參考文獻.........................................53
附錄.................................................60

 
表目錄
表3.1 淹水區之水面及空地於不同偏極上的反射情形對照表.14
表4.2 本研究使用之SAR影像系統參數....................18
表4.3 誤差矩陣.......................................27
表4.4 類神經網路淹水辨識模式輸入變數組合.............29
表4.5模式一不同輸入變數組合之辨識結果................33
表4.6 模式一最佳輸入變數組合之誤差矩陣...............34
表4.7模式二不同輸入變數組合之辨識結果................36
表4.8 模式二最佳輸入變數組合之誤差矩陣...............37
表4.9模式三不同淹水門檻值之辨識結果..................39
表4.10 模式三最佳淹水門檻值之誤差矩陣................40
表4.11 各模式訓練與測試之結果比較表..................42
表4.12 各模式訓練及測試之誤差矩陣比較表..............43
表4.13 各模式訓練與測試之結果比較表..................46
附表1 ALOS衛星SAR影像拍攝方式與感測器規格............60
附表2 ALOS衛星SAR影像各等級產品規格表................62
 
圖目錄
圖3.1 BPNN架構圖.....................................8
圖3.2 合成孔徑雷達意識圖.............................13
圖3.3 SAR影像與地表物體之反射情形....................15
圖4.1得子口溪水系地理位置與研究區域相對位置圖........17
圖4.2淹水前SAR衛星影像（a）HH頻道，（b）HV頻道.......19
圖4.3淹水時SAR衛星影像...............................20
圖4.4 縣府調查淹水區.................................22
圖4.5 現地調查淹水區.................................23
圖4.6 模式所使用之淹水區及非淹水區範本...............24
圖4.7 訓練與測試區之分割圖...........................25
圖4.8 模式一架構圖...................................30
圖4.9 模式二架構圖...................................31
圖4.10 模式一淹水辨識結果............................35
圖4.11 模式二淹水辨識結果............................38
圖4.12 模式三淹水辨識結果............................41
圖4.13 模式一與模式二辨識結果之差異..................43
圖4.14淹水前與淹水時SAR影像差異分布圖................44
圖4.15 模式一修正後淹水辨識結果......................46
圖4.16 模式二修正後淹水辨識結果......................47
圖4.17 模式三修正後淹水辨識結果......................47
圖4.18 壯圍鄉新社村及古亭村之可見光影像原圖..........49
圖4.19 模式二於壯圍鄉新社村及古亭村之淹水區域推估....49
圖4.20 三星鄉尾塹村之可見光影像圖....................50
圖4.21模式二於三星鄉尾塹村之淹水區域推估結果.........50
附圖1 ALOS衛星SAR影像拍攝方式意識圖..................61

1.Bach, H., Appel, F., Fellah, K. and de Fraipont, P. (2005). Application of flood monitoring from satellite for insurances. Paper presented at the IEEE Geoscience and Remote Sensing Society.
2.Bonansea, E. (1995) Mapping of 1994 flood in Piedmonte region: an example of remote sensing and GIS application, First ERS Thematic Working Group Meeting on Flood Monitoring, Frascati (Roma, Italy), 26-27 June 1995 .
3.Brivio, P. A., Colombo, R., Maggi, M. and Tomasoni, R. (2002). Integration of remote sensing data and GIS for accurate mapping of flooded areas. International Journal of Remote Sensing, 23(3), 429-441.
4.Chang, L.-C., Shen, H.-Y., Wang, Y.-F., Huang, J.-Y. and Lin, Y.-T. (2010). Clustering-based hybrid inundation model for forecasting flood inundation depths. Journal of Hydrology, 385(1-4), 257-268.
5.Chen, C.-S., Chen, B. P.-T., Chou, F. N.-F. and Yang, C.-C. (2010). Development and application of a decision group back-propagation neural network for flood forecasting. Journal of Hydrology, 385(1-4), 173-182.
6.Conger, C. L., Fletcher, C. H. and Barbee, M. (2005). Artificial neural network classification of sand in all visible submarine and subaerial regions of a digital image. Journal of Coastal Research, 21(6), 1173- 1177.
7.Coulibaly, L., Migolet, P., Adegbidi, H. G., Fournier, R. and Hervet, E. (2008). Mapping aboveground forest biomass from IKONOS satellite image and multi-source geospatial data using neural networks and a kriging interpolation. Paper presented at the IEEE Geoscience and Remote Sensing Society. 
8.Dey, C., Jia, X. and Fraser, D. (2008). Decision fusion for reliable flood mapping using remote sensing images. 184-190.
9.Baldassarre, G., Schumann, G. and Bates, P. D. (2009A). Near real time satellite imagery to support and verify timely flood modelling. Hydrological Processes, 23(5), 799-803.
10.Baldassarre, G., Schumann, G. and Bates, P. D. (2009B). A technique for the calibration of hydraulic models using uncertain satellite observations of flood extent. Journal of Hydrology, 367(3-4), 276-282.
11.Foody, C. M. (2002). Status of land cover classification accuracy assessment. Remote Sensing of Environment, 80, 185- 201.
12.Freeman, A. (1992). SAR calibration an overview. IEEE Geoscience and Remote Sensing Society, 30(6), 1107- 1121.
13.Galloway, D. L. and Hoffmann, J. (2006). The application of satellite differential SAR interferometry-derived ground displacements in hydrogeology. Hydrogeology Journal, 15(1), 133-154.
14.Henry, J. B., Chastanet, P., Fellah, K. and Desnos, Y. L. (2003). ENVISAT multipolarised ASAR data for flood mapping. IEEE Geoscience and Remote Sensing Society, 7, 1136- 1138 vol.1132.
15.Hoffmann, J. (2005). The future of satellite remote sensing in hydrogeology. Hydrogeology Journal, 13(1), 247-250.
16.Hoffmann, J. and Sander, P. (2006). Remote sensing and GIS in hydrogeology. Hydrogeology Journal, 15(1), 1-3.
17.Horritt, M. (2006). A linearized approach to flow resistance uncertainty in a 2-D finite volume model of flood flow. Journal of Hydrology, 316(1-4), 13-27.
18.Hossain, A. K. M. A., Chao, X. and Jia, Y. (2009). Mapping 2008 Midwest flood dynamics using multi-sensor  remotely sensed data, a case study of CCHE2D Flood Simulation  Results Validation. Paper presented at the MAESC 2009 "Safety: A Must - Not an Option". 
19.Hostache, R., Lai, X., Monnier, J. and Puech, C. (2010). Assimilation of spatially distributed water levels into a shallow-water flood model. Part II: Use of a remote sensing image of Mosel River. Journal of Hydrology, 390(3-4), 257-268.
20.Hostache, R., Matgen, P., Schumann, G., Puech, C., Hoffmann, L. and Pfister, L. (2009). Water level estimation and eduction of hydraulic model calibration uncertainties using satellite SAR images of Floods. IEEE Geoscience and Remote Sensing Society, 47(2), 431- 441.
21.Jun, J.-N., Seo, D.-C. and Lim, H.-S. (2004). Calculation of agricultural land flooding disaster area by typhoon from KOMPSAT-1 EOC satellite image data. Paper presented at the IEEE Geoscience and Remote Sensing Society. 
22.Juneja, M., Walia, E., Sandhu, P. S. and Mohana, R. (2009). Implementation and comparative analysis of rough set, Artificial neural network (ANN) and fuzzy rough classifiers for satellite image classification. Paper presented at the Intelligent Agent & Multi-Agent Systems. 
23.Kawano, N. and Shimada, M. (2009). Flood disaster monitoring with ALOS PALSAR observation. IEEE Geoscience and Remote Sensing Society, 5, IV-641- IV644.
24.Khashman, A. and Dimililer, K. (2008). Image compression using Neural Networks and Haar Wavelet. WSEAS Transactions on Signal Processing, 4(5), 330- 339.
25.Klobucar, D., Pernar, R., Loncaric, S. and Subasic, M. (2008). Artificial neural networks in the assessment of stand parameters from an IKONOS satellite image. Croatian Journal of Forest Engineering, 29(2), 201- 211.
26.Lai, X. and Monnier, J. (2009). Assimilation of spatially distributed water levels into a shallow-water flood model. part I: mathematical method and test case. Journal of Hydrology, 377(1-2), 1-11.
27.Lazaro, J., Arias, J., Martin, J., Zuloaga, A. and Cuadrado, C. (2006). SOM segmentation of gray scale images for optical recognition. Pattern Recognition Letters, 27(16), 1991-1997.
28.Leon, J., Calmant, S., Seyler, F., Bonnet, M., Cauhope, M., Frappart, F., et al. (2006). Rating curves and estimation of average water depth at the upper Negro River based on satellite altimeter data and modeled discharges. Journal of Hydrology, 328(3-4), 481-496.
29.Lin, C. A., Wen, L., Lu, G., Wu, Z., Zhang, J., Yang, Y., et al. (2010). Real-time forecast of the 2005 and 2007 summer severe floods in the Huaihe River Basin of China. Journal of Hydrology, 381(1-2), 33-41.
30.Mandar, S. and Rashmi, S. (2011). Textural feature based image classification using artificial neural network. Communications in Computer and Information Science, 125, 62- 69.
31.Marique, T., Kharoubi, A., Bauffe, P. and C., D. (2003). Modeling of fried potato chips color classification using image analysis and artificial neural network. Food Engineering and Physical Properties, 68(7), 2263- 2266.
32.Mason, D. C., Bates, P. D. and Dall’ Amico, J. T. (2009). Calibration of uncertain flood inundation models using remotely sensed water levels. Journal of Hydrology, 368(1-4), 224-236.
33.Mendoza, O., Melin, P. and Licea, G. (2009). A hybrid approach for image recognition combining type-2 fuzzy logic, modular neural networks and the Sugeno integral. Information Sciences, 179(13), 2078-2101.
34.Moussa, R. and Bocquillon, C. (2009). On the use of the diffusive wave for modelling extreme flood events with overbank flow in the floodplain. Journal of Hydrology, 374(1-2), 116-135.
35.Profeti, G. and MaciNtonsh, H. (1997). Flood management through Landsat TM and ERS SAR data: a case study. Hydrological  Processes, 11, 1397- 1408.
36.Rasmi, S. and Mandar, S. (2011). Textural feature based image classification using artificial neural network. Communications in Computer and Information Science 2011 , 125, Part 1 62-69.
37.Robins, N. S., Davies, J., Farr, J. L. and Calow, R. C. (2006). The changing role of hydrogeology in semi-arid southern and eastern Africa. Hydrogeology Journal, 14(8), 1483-1492.
38.Schumann, G., Matgen, P., Hoffmann, L., Hostache, R., Pappenberger, F. and Pfister, L. (2007). Deriving distributed roughness values from satellite radar data for flood inundation modelling. Journal of Hydrology, 344(1-2), 96-111.
39.Shah, P. S. K. and Gandhi, V. (2004). Image classification cased on textural features using Artificial Neural Network (ANN). Januar IE(I) Journal– ET, 84, 72- 77.
40.Shimada, M., Isoguchi, O., Tadono, T. and Isono, K. (2009). PALSAR radiometric and geometric calibration. IEEE Geoscience and Remote Sensing Society, 47(12), 3915- 3932.
41.Singh, V. and Mishra, R. (2006). Developing a machine vision system for spangle classification using image processing and artificial neural network. Surface and Coatings Technology, 201(6), 2813-2817.
42.Sorooshian, S., Hsu, K., Imam, B. and Hong, Y. (2007). Global precipitation estimation from satellite image using artificial neural networks. In H. Wheater (Ed.), Hydrological Modelling in Arid and Semi-Arid Areas (pp. 21- 28).
43.Srivastav, S. K., Lubczynski, M. W. and Biyani, A. K. (2007). Upscaling of transmissivity, derived from specific capacity: a hydrogeomorphological approach applied to the Doon Valley aquifer system in India. Hydrogeology Journal, 15(7), 1251-1264.
44.Stathakis, D. and Vasilakos, A. (2006). Satellite image classification using granular neural networks. International Journal of Remote Sensing, 27(18), 3991-4003.
45.Tao, X. and Michel, H. E. (2003). Classification of multispectral satellite image data using improved NRBF neural networks. Paper presented at the Intelligent Robots and Computer Vision XXI: Algorithms, Techniques, and Active Vision. 
46.Tralli, D., Blom, R., Zlotnicki, V., Donnellan, A. and Evans, D. (2005). Satellite remote sensing of earthquake, volcano, flood, landslide and coastal inundation hazards. ISPRS Journal of Photogrammetry and Remote Sensing, 59(4), 185-198.
47.Tuan, T. A., Duong, N. D. and Vietnam. (2009). Flood monitoring using ALOS PALSAR imagery. Paper presented at the 7th FIG Regional Conference Spatial Data Serving People: Land Governance and the Environment – Building the Capacity. 
48.Turkar, V. (2010). Polarimetric SAR image classification by using artificial neural network. Paper presented at the International Conference and Workshop on Emerging Trends in Technology. 
49.Wang, C.-T., Chen, K.-S., Lee, H.-W., Lee, J.-S., Boerner, W.-M., Wang, R.-Y., et al. (2007). Disaster monitoring and environmental alert in taiwan by repeat-pass spaceborne SAR. Paper presented at the Geoscience and Remote Sensing Symposium. 
50.Yamada, Y. (2003). Relation between ground features and mathematical morphology using JERS-1 SAR data during flooding time in paddy areas. Paper presented at the IEEE Geoscience and Remote Sensing Symposium. 
51.Yamada, Y. (2007). Flood mapping for paddy areas in Thailand and Indonesia using PALSAR data. Paper presented at the Pro. DVD Of the first joint PI symposium of ALOS Data Nodes for ALOS science program. 
52.Yamada, Y. (2008). Morphological analysis of flood inundated regions in paddy areas using los/palsar data and its distribution on the google earth design of the future disaster management system (fdms). International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 36(B4).
53.Youngjoo, K., Akira, H., Hironori, I. and Kazuhiko, F. (2011). Flood risk assessment using inundation depth model and ALOS images: A Case Study In Kabul River, Pakisatn. Paper presented at the 2011 European Geosciences Union General Assembly. 
54.Zeeb, C., Göckus,D., Bons, P., Ajmi, H., Rausch, R. and Blum, P. (2010). Fracture flow modelling based on satellite images of the Wajid Sandstone, Saudi
Arabia. Hydrogeology Journal, 18(7), 1699-1712.
55. Zyl, J. J. v. (1990). Calibration of polarimetric radar images using only image
parameters and trihedral corner reflector responses. IEEE Geoscience and
Remote Sensing Society, 28(3), 337-348.
56. 江介倫，鄭克聲，陳錕山，2005，應用SPOT 及空載全偏極合成孔徑雷達
影像判識地表覆蓋，農業工程學報51(4),84- 96.
57. 林文賜，周天穎，林昭遠，2001，應用監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6(1),41- 58.
58. 張中白，2005) ，大地彩虹合成孔徑雷達. 科學發展390, 18- 23.
59. 張斐章，張麗秋，2010，類神經網路導論：原理與應用，滄海書局
60. 陳錕山，1995，遙測影像之整合研究2-子計畫4,合成孔徑雷達影像輻射校
正與影像分析，行政院國家科學委員會微縮小組
61. 曾煥君，王志添，許明光，陳錕山，2003，合成口徑雷達衛星影像應用於
颱風時河道狀態之監測，航測及遙測學刊， 8(4), 83- 98.
62. 楊孟學，林明璋，劉進金，2009，結合衛星影像與地形指標於山崩自動分
類之研究. 航測及遙測學刊, 14(1), 11- 23.
63. 楊龍士，雷祖強，周天穎，2006，遙感探測理論與分析實務，文魁資訊公
司
64. 蔡孟蓉，江衍銘，張斐章，2007，類神經網路結合衛星影像於颱風降雨預
報，臺灣水利, 55(1), 22- 31.
65. 潘國樑，2009，遙測學大綱，科技圖書公司