本研究是利用全偏極合成孔徑雷達(fully polarimetric SAR)資料於目標分類。全偏極合成孔徑雷達特徵包含於偏極共變異矩陣(polarimetric covariance matrix)內，矩陣中除線性偏極特徵外，尚包含了偏極間之交互作用項。以全偏極雷達特徵作為分類資訊相較於已往以部分偏極特徵為分類資訊，其好處在於保有更完整之地物散射特性。
利用全偏極雷達特徵於地表分類時，由於一般所使用之Euclidean距離並不適用於全偏極雷達資料之分佈特性，因此本研究所用之距離估算方法為架構於複數高斯分佈(complex Gaussian distribution)所計算出之統計距離。另外，本研究所利用之模糊類神經網路，其優點在於不需預先了解各類別之中心及分佈，僅需挑選少許訓練區供類神經網路訓練；類神經網路會根據訓練區之特徵及模糊隸屬度以疊代方式產生對應，並於完成訓練後對整組資料進行分類。本模糊類神經網路所採用之模糊演算法為fuzzy c-means演算法。
於測試方面，所用的資料為鰲鼓農場之空載全偏極合成孔徑雷達資料。由於合成孔徑雷達資料皆有斑駁(speckle)現象，斑駁會降低資料之訊雜比，致使分類正確率降低；因此，在應用全偏極合成孔徑雷達資料於地表分類前，需先經過斑駁抑減(speckle reduction)以降低雜訊之影響。最後，測試的結果顯示以全偏極資料進行地表分類時，複數高斯距離能有效地估計全偏極資料與類別中心之距離及模糊隸屬度，並於神經網路進行分類時有良好的分類結果。
另外，結合類神經網路及fuzzy c-means演算法，改進了fuzzy c-means非監督式分類之缺點，合併成為一監督式之模糊神經網路分類器。

This paper will present a method based on a fuzzy neural network that will use fully polarimetric information for SAR image classification. The approach makes use of the statistical properties of the polarimetric data while taking advantage of a fuzzy neural network that requires no a priori information about the data. A distance measure based on the complex Gaussian distribution was applied to the fuzzy clustering algorithm and then subsequently incorporated into the neural network. Instead of pre-selecting the polarization channels as has usually been done before, the inputs to the neural network are now all elements of the covariance matrix which serve as the target feature vector. It is thus expected that the neural network will be able to take full power of the fully polarimetric information for the purposes of image classification. With the generalization, adaptation, and other capabilities of the neural network, general information contained in the covariance matrix, such as the amplitude, phase difference, degree of polarization, etc. are well preserved and thus are fully explored. One of the essential features in this setup lies in that the chosen neural network must be able to handle such high dimensional and yet diverse input feature vectors, while maintaining a sufficiently fast learning speed in order drive itself as a practical tool. To demonstrate the advantages of the proposed method, we compare four different configurations, which are categorized by their uses of feature vectors, classifier, distance measures, and whether fuzzy c-means are applied or are applicable. The validity and effectiveness of the proposed scheme support the utilization of this polarimetric information. It is shown that with fully polarimetric data, the fuzzy neural network can substantially reduce the learning time and improve the classification accuracy as well. It must be noted that a Lee polarimetric filter, that reduces the speckle noise while preserving the polarimetric properties has proven to be useful in improving the classification accuracy. It is also demonstrated that the proposed approach gains adaptability and flexibility for high dimensional feature vectors, such as the complete polarimetric data.

[1] C.H. Chen, ed.,Information Processing for Remote Sensing, World Scientific, 1999.
[2] M. R. Azimi-Sadjadi, S. Ghaloum, and R. Zoughi, “Terrain classification in SAR images using principal components analysis and neural network,” IEEE Transactions on Geoscience and Remote Sensing, vol. 31, no. 2, pp. 511-515, Mar. 1993.
[3] Y. Hara, R. G. Atkins, S. H. Yueh, R. T Shin, and J. A. Kong, “Application of neural networks to radar image classification,” IEEE Transactions on Geoscience and Remote Sensing, vol. 32, no. 1, pp. 100-109, Jan. 1994.
[4] G. A. Carpenter and S. A. Grossberg, “The ART of adaptive pattern recognition by self-organizing neural network,” IEEE Comput., vol. 21, pp. 77-88, Mar. 1988.
[5] T. Kohonen, Self-Organization and Associative Memory. Berlin: Springer-Verlag, 1989.
[6] J. S. Lee, M. R. Grunes, and G. de Grandi, “Polarimetric SAR speckle filtering and its implication for classification,” IEEE Transactions on Geoscience and Remote Sensing, vol. 37, no. 5, pp. 2363-2373, Sep. 1999.
[7] G. M. Foody, M. B. McCulloch, and W. B. Yates, “Crop classification from C-band polarimetric radar data,” Int. J. Remote Sensing, vol. 15, no. 14, pp. 2871-2885, 1994.
[8] K. S. Chen, W.P. Huang, D.W. Tsay and F. Amar, “Classification of multifrequency polarimetric SAR image using a dynamic learning neural network,” IEEE Transactions on Geoscience and Remote Sensing, vol. 34, no. 3, pp. 814-820, May 1996.
[9] P. D. Heermann and N. Khazene, “Classification of multispectral remote sensing data using a back-propagation neural network,” IEEE Transactions on Geoscience and Remote Sensing, vol. 30, no. 1, pp.81-88, Jan. 1992.
[10] H. Bischof, W. Schneider, and A. H. Pinz, “Multispectral classification of Landsat-image using neural network,” IEEE Transactions on Geoscience and Remote Sensing, vol. 30, no. 2,pp.482-490, May 1992.
[11] K. S. Chen, Y. C. Tzeng, C. F. Chen, and W. L. Kao, “Land-cover classification of multispectral imagery using a dynamic learning neural network,” Photogramm. Eng. Remote Sens., vol. 61, no. 4, Apr. 1995.
[12] J. A. Benediktsson, P. H. Swain, and O. K. Ersoy, “Neural network approaches versus statistical methods in classification of multisource remote sensing data,” IEEE Transactions on Geoscience and Remote Sensing, vol. 28, no.4, pp.540-552, July 1990.
[13] L. Du, J. S. Lee, and S. A. Mango, “Fuzzy classification of earth terrain covers using multi-look polarimetric SAR image data,” in Proc. 1993 Int. Geoscience and Remote Sensing Symp. (IGARSS’ 93), vol. 4, pp. 1602-1604, 1993.
[14] Y. C. Tzeng and K. S. Chen, “A fuzzy neural network to SAR image classification,” IEEE Transactions on Geoscience and Remote Sensing, vol. 36, no. 1, pp.301-307, Jan. 1998.
[15] J. S. Lee, K. W. Hoppel, S. M. Mango, and A. R. Miller, “Intensity and phase statistics of multilook polarimetric and interferometric SAR imagery,” IEEE Trans. Geosci. and Remote Sensing, vol.32, no.5, pp.1017-1028, Sep. 1994.
[16] J. J. van Zyl, H. A. Zebker, and C. Elachi, “Imaging radar polarimetric signatures: Theory and observation,” Radio Sci., vol. 22, pp. 529-543, 1987.
[17] V. B. Tayloer, “CYLOPS: The JPL AIRSAR synoptic processor,” in Proc. 1992 Int. Geosci. Remote Sensing Symp. (IGARSS’92), Houston. TX, 1992, pp. 652-654.
[18] N. R. Goodman, “Statistical analysis based on a certain complex Gaussian distribution (an introduction),” Ann. Mathemat. Statist., vol. 34, pp. 152-177, 1963.
[19] S. Goze and A. Lopes, “A MMSE speckle filter for full resolution SAR polarimetric data,” J. Electron. Waves Applicat., vol. 7, no. 5, pp. 717-737, 1993.
[20] A. Lopes and F. Sery, “Optimal speckle reduction for the product model in multi-look polarimetric SAR imagery and the Wishart distribution,” IEEE Transactions on Geoscience and Remote Sensing, vol. 35, no. 3, pp. 632-647, May 1997.
[21] J. C. Bezdek, Pattern Recognition with Fuzzy Objective Function Algorithms, New York: Plenum Press, 1987.
[22] J. S. Lee, M. R. Grunes, and R. Kwok, “Classification of multi-look polarimetric SAR imagery based on complex Wishart distribution,” Int. J. Remote Sensing, vol. 15, no. 11, pp. 2299-2311, 1994.
[23] F. Wang, “Fuzzy supervised classification of remote sensing images,” IEEE Transactions on Geoscience and Remote Sensing, vol. 28, no. 2, pp. 194-201, Mar. 1990.
[24] L. Du and J. S. Lee, “Fuzzy classification of earth terrain covers using complex polarimetric SAR data,” Int. J. Remote Sensing, vol. 17, no. 4, pp. 809-826, 1996.
[25] F. T. Ulaby and C. Elachi, Radar Polarimetry for Geoscience Applications, Norwood: Artech HOUSE, 1990
[26] J. S. Lee, “Refined filtering of image noise using local statistics,” Computer Graphics and Image Processing, vol. 15, pp. 3880-389, 1981.
[27] J. S. Lee, M. R. Grrunes, and S. A. Mango, “Speckle reduction in multipolarization multifrequency SAR imagery,” IEEE Transaction on Geoscience and Remote Sensing, vol. 29, no. 4, pp. 535-544, July 1991.
[28] Y. C. Tzeng, K. S. Chen and C. T. Chen, Integration of SPOT and SAR images for monitoring of environmental changes by a Fuzzy Neural Network, Remote Sensing Applications in Geoscience.