臺灣博碩士論文加值系統

多光譜遙測影像已被廣泛應用到地表覆蓋之分類的領域上，監督式分類法以所選取各類別中具代表性的訓練樣本(training sample)之分類特徵值為基礎進行統計分析，故在影像分類中會有因為訓練樣本不同而產生的不確定性(uncertainty)。在遙測領域中一般多將分類結果以混淆矩陣(confusion matrix)的形式評估其分類的準確度(accuracy)，然而訓練樣本的不確定性使得混淆矩陣中的整體準確度、各類別之使用者準確度和生產者準確度也存在著不確定性，需由多組之分類結果對取樣所造成之不確定性做進一步的分析及探討。除此之外，高斯最大概似分類法(Gaussian maximum likelihood classifier)中假設各類別之樣本分類特徵值呈多變量常態分布(multivariate normal distribution)。然而實際上許多類別之地物分類特徵值並不符合此假設，此情況會影響分類結果的準確度。
本研究利用ALOS衛星所提供之多光譜遙測影像對台北市及其近郊地區做地表覆蓋之分類。第一部分中，利用常態分布轉換函數對原始樣本之分類特徵值做轉換，再以經函數轉換後之樣本進行影像分類，欲藉此改善分類結果之準確度。第二部分中，使用跋靴法之重複取樣(bootstrap resampling)，由原始訓練樣本組中取樣以產生多組的訓練樣本，並以取樣所得之訓練樣本各別進行影像分類，模擬實際中情況因為取樣不同所產生之不同的分類結果。將多組分類結果準確度之分布特性做分析，可評估分類準確度中之不確定性；將多組分類結果統計整合，並設定門檻值將在分類結果中具有較大不確定性之像元指定為「未分類」(unclassified)類別後，以包含「未分類」類別之分類地圖呈現，可分析含分類之不確定性在空間上的分布資訊。
研究結果顯示，若在分類前將樣本特徵值經函數轉換，使得分類所使用之各類別樣本分類特徵值符合多變量常態分布之假設，可改善分類結果之準確度。藉由跋靴法之重複取樣所產生之多組分類結果準確度及多組分類結果地圖，可發現在分類準確度的評估中，準確度較差之類別存在有較大的不確定性，且因取樣所造成的不確定性主要是由包含混合類別之像元所造成，故應用跋靴法之重複取樣於影像分類中有助於評估其中因為取樣不同所造成的不確定性。

Multispectral remote sensing images are widely used for landuse/landcover (LULC) classification. Performance of such classification practices is normally evaluated through the confusion matrix which summarizes the producer’s and use’s accuracies and the overall accuracy. However, the confusion matrix is based on the classification results of a set of multi-class training data. As a result, the classification accuracies are heavily dependent on the representativeness of the training data set. It is imperative for practitioners to assess the uncertainties of LULC classification in order to obtain a full understanding of the classification results. In addition, the Gaussian-based maximum likelihood classifier (GMLC) is widely applied in many practices of LULC classification. The GMLC assumes the classification features jointly form a multivariate normal distribution, whereas, in reality, may features of individual landcover classes have been found to be non-Gaussian. Direct application of GMLC will certainly affect the classification results.
In the study conducted in Taipei and its vicinity, the satellite images acquired by the AVNIR-2 sensor onboard the ALOS satellite were used. We tackled those two problems by firstly transforming the original training data set to a corresponding data set which forms a multivariate normal distribution before conducting classification using GMLC. Then, we applied the bootstrap resampling technique to generate a large set of multi-class resampled training data set from the original training data set. LULC classification was the implemented for each resampled training data set using GMLC. Finally, the uncertainties of LULC classification accuracies were assessed by evaluating the distributions of the accuracies derived from a set of confusion matrices. Combining the resampled bootstrap results of classification for each pixel and setting a threshold, pixels with higher uncertainties would be assigned to “unclassified”. The spatial characteristics of the uncertainties of LULC classification were assessed by showing the location of the unclassified pixels in the map.
Results of this study demonstrate that Gaussian-transformation of the original training data achieved better classification accuracies, and that the bootstrap resampling technique is a very helpful tool for assessing uncertainties of LULC classification because it could assess the uncertainties in classification accuracies and illustrate the mixed pixels in the study area.

摘要... I
ABSTRACT... III
目錄... V
表目錄... VII
圖目錄... VIII
第一章 前言... 1
1.1 研究動機及目的... 1
1.2 研究架構及流程 ... 2
第二章 文獻回顧... 5
2.1 遙測影像分類... 5
2.2 遙測影像分類中之不確定性探討... 7
第三章 理論與方法... 11
3.1 遙測影像分類... 11
3.1.1 高斯最大概似法(Gaussian Maximum Likelihood Classifier)... 12
3.1.2 貝氏分類法(Baysian Classifier)... 14
3.1.3 建立分類結果之門檻值... 16
3.2 分類結果準確度之評估... 18
3.3 常態分布函數轉換... 19
3.4 跋靴法... 21
第四章 研究區域與研究資料... 25
4.1 研究區域... 25
4.2 資料選取... 25
4.2.1 衛星影像... 25
4.2.2 訓練樣本之選取... 26
第五章 結果與討論... 35
5.1 函數分布轉換對分類結果準確度之影響... 35
5.2 以跋靴法重複取樣之樣本做分類 ... 41
5.2.1 分類準確度之不確定性分析... 41
5.2.2 全幅影像之不確定性分析... 47
5.3 「未分類」類別之分析... 61
5.3.1 「未分類」類別像元之位置分布... 61
5.3.2 「未分類」類別之分類特徵值... 69
第六章 結論與建議... 75
6.1 結論... 75
6.2 建議... 77
參考文獻... 79

1. 朱子豪，1980，遙測土地利用調查系統先驅計畫。
2. 江介倫、鄭克聲，2004，類神經網路分類法之探討及其於衛星影像分類之應用，農業工程學報，50.1, 24-34
3. 周心怡，2004，拔靴法 (Bootstrap) 之探討及其應用，國立中央大學統計研究所碩士論文。
4. 洪維均，2010，遙測技術應用於不同區域都市發展程度之監測，國立臺灣大學生物環境系統工程學研究所博士論文。
5. 連以婷，2010，水文模式之參數不確定性分析，國立臺灣大學生物環境系統工程學研究所碩士論文。
6. 黄恩興，2010，遥感影像分類结果的不確定性研究，中國農學通報，26.5, 322-325
7. 鍾建文，2006，階段式遙測影像分類應用於土地利用變遷偵測之研究，國立臺灣大學生物環境系統工程學研究所碩士論文。
8. Albeanu, G., 2007. On using bootstrap approach for uncertainty estimation. In G. Guslicov (ed.), Proceedings of the 1st International Proficiency Testing Conference,221-229.
9. Asmala, A., 2012, Analysis of Maximum Likelihood Classification on Multispectral Data. Applied Mathematical Sciences, 6.129-132: 6425-6436.
10. Bo, Y. C., & Wang, J. F., 2008. A general method for assessing the uncertainty in classified remotely sensed data at pixel scale. Spatial Uncertainty, Proceedings of Accuracy, 186-194.
11. Congalton, R. G., 1991, A review of assessing the accuracy of classifications of remotely sensed data. Remote sensing of environment, 37.1: 35-46.
12. Deng, J.S., Wang, K., Hong, Y., Qi, J.G., 2009. Spatio-temporal dynamics and evolution of land use change and landscape pattern in response to rapid urbanization. Landscape and Urban Planning, 92, 187–198.
13. Efron, B., 1979. Bootstrap Methods：Another Look at the Jackknife. The Annals of Statistics, 7, 1-26.
14. Efron, B., Tibshirani, R.J., 1993. An Introduction to the Bootstrap. Chapman & Hall.
15. Firestone, Michael, et al., 1997, Guiding principles for Monte Carlo analysis. Risk Assessment Forum, US Environmental Protection Agency.
16. Foody, G. M. and Atkinson, P. M. (Eds.)., 2002. Uncertainty in Remote Sensing and GIS. J. Wiley.
17. Goncalves, L. M., Fonte, C. C., Julio, E. N., & Caetano, M., 2009. A method to incorporate uncertainty in the classification of remote sensing images. International Journal of Remote Sensing, 30(20), 5489-5503.
18. Hammond, T. O. and Verbyla, D. L., 1996. Optimistic bias in classification accuracy assessment. International Journal of Remote Sensing, 17(6), 1261-1266.
19. Hung, W.C, Chen, Y.C and Cheng, K.C, 2010. Comparing landcover patterns in Tokyo, Kyoto, and Taipei using ALOS multispectral images. Landscape and Urban Planning, 97.2, 132 -145
20. Ji, W., Ma, J., Twibell, R.W and Underhill, K., 2006. Characterizing urban sprawl using multi-stage remote sensing images and landscape metrics. Computers, Environment and Urban Systems, 30, 861–879.
21. Lillesand, T.M. and Kiefer, R.W., 2000. Remote Sensing and Image Interpretation. N.Y., John Wiley & Sons.
22. Luck, M. and Wu, J., 2002. A gradient analysis of urban landscape pattern: a case study from the Phoenix metropolitan region, Arizona, USA. Landscape Ecology, 17, 327–339.
23. McIver, D.K. and Friedl, M.A., 2002. Using prior probabilities in decision-tree classification of remotely sensed data. Remote Sensing of Environment.81, 253– 261.
24. Nishii, R. and Morisaki, Y., 2001. Spatial discriminant analysis based on power-elliptic distributions and power transformations. In Geoscience and Remote Sensing Symposium, 2001. IGARSS''01. IEEE 2001 International, Vol. 7, 2991-2993.
25. Okeke, F., & Karnieli, A., 2006. Methods for fuzzy classification and accuracy assessment of historical aerial photographs for vegetation change analyses. Part I: Algorithm development. International journal of remote sensing, 27(1), 153-176
26. Ress, W.G., 1990. Physical Principle of remote sensing, Combridge University Press.
27. Strahler, A. H., 1980, The use of prior probabilities in maximum likelihood classification of remotely sensed data. Remote Sensing of Environment, 10.2, 135-163.
28. Verbyla, D. L. and Hammond, T. O., 1995. Conservative bias in classification accuracy assessment due to pixel-by-pixel comparison of classified images with reference grids. Remote Sensing, 16(3), 581-587.
29. Weber, K.T., and J. Langille, 2007. Improving classification accuracy assessments with statistical bootstrap resampling techniques. GIS Science and Remote Sensing, 44 (3), 237-250