隨著遙測技術的進步，近年來高光譜影像的應用越來越廣泛，由於影像中每個像素都包含了上百個波段的資訊，要從如此巨量的資料中，找尋影像中的異常物，是一項艱難的挑戰。在之前的研究中，知名的RX演算法與高次方統計法已被運用來解決這個問題。然而RX演算法雖然可以偵測出異常物，卻無法區分這些異常物的種類；高次方統計法雖可彌補無法辨別各類異常物的缺點，但必須重複找尋新的投影方向，需要花費較長的時間。在這篇論文中，我們根據RX演算法，提出一個新的能有效偵測且可區分異常物類別的偵測演算法，稱為「背景白化目標物偵測演算法」(BWTDA)。我們假設影像中的背景物為高斯分布並先將資料白化，經過白化的過程後，各個維度(光譜波段)上的資料成為彼此獨立且單一的分布型態(independent-identical-distributed)，接著利用「目標物偵測程序」(Target Detection Process)，自動的找到可能的異常物，最後透過「目標物分類程序」(Target Classification Process)，區分出異常物的種類。實驗的結果顯示我們設計的方法改善了RX演算法無法區分異常物種類的缺點，且比高次方統計的方法更為快速。

With the improvement of remote sensing technology, applications of hyperspectral imaging have rapidly grown in various fields. Hyperspectral imaging collects information of the surface with hundreds of channels resulting in hundreds of co-registered images. To process such large amount of data without prior information of the scene is challenging, especially for anomaly detection. Several existing methods are devoted to this problem, such as the well-known RX algorithm and high-moment statistics approaches. The RX algorithm can detect all anomalies in a single image but it cannot classify them. On the other hand, the high-moment statistics approaches use parameters such as skewness and kurtosis to find the projection directions recursively, which is computationally expensive and time-consuming. This dissertation proposes an anomaly detection and classification algorithm extended from the RX algorithm named the Background Whitened Target Detection Algorithm. It first performs background whitening by assuming a Gaussian distribution. After the whitening, the background will be independent-identical-distributed Gaussian in all spectral bands. The Target Detection Process then searches for potential anomalies automatically, and the Target Classification Process classifies them individually. Experimental results show that the proposed method can improve upon the RX algorithm by classifying the anomalies and outperforming the original high-moment statistics approach in computational time.

[1]"AN OVERVIEW OF HYPERSPECTRAL REMOTE SENSING," CSIRO Earth Observation Centre, [Online]. Available: http://www.eoc.csiro.au/hswww/Overview.htm.
[2]Shaw, G. A. and H. K. Burke, "Spectral imaging for remote sensing," Lincoln laboratory journal, vol. 14, no. 1, pp. 3-28, 2003.
[3]Reed, I. S. and X. Yu, "Adaptive multiple-band CFAR detection of an optical pattern with unknown spectral distribution," IEEE Trans. on Acoustic, Speech and Signal Processing, vol. 38, pp. 1760-1770, 1990.
[4]Yu, X., I. S. Reed and A. D. Stocker, "Comparative performance analysis of adaptive multispectral detectors," IEEE Trans. Signal Process, vol. 41, pp. 2639-2656, 1993.
[5]E. A. Ashton, "Detection of subpixel anomalies in multispectral infrared imagery using an adaptive Bayesian classifier," IEEE Trans. on Geoscience and Remote Sensing, vol. 36, pp. 506-517, 1998.
[6]Ifarraguerri, A. and C.-I. Chang, "Unsupervised hyperspectral image analysis with projection pursuit," IEEE Trans. on Geoscience and Remote Sensing, vol. 38, pp. 2529-2538, 2000.
[7]Schweizer, M. and J. M. F. Moura, "Efficient detection in hyperspectral imagery," IEEE Trans. on Image Processing, vol. 10, pp. 584-597, 2001.
[8]Chiang, S. S., C.-I. Chang and I. W. Ginsberg, "Unsupervised target detection in hyperspectral images using projection pursuit," IEEE Trans. on Geoscience and Remote Sensing, vol. 39, pp. 1380-1391, 2001.
[9]Chiang, S. S., C.-I. Chang and I. W. Ginsberg, "Unsupervised hyperspectral image analysis using independent components analysis," in IEEE 2000 International Geoscience and Remote Sensing Symposium, Hawaii, 2000.
[10]Ren, H., Q. Du, J. Wang, C.-I. Chang, J. O. Jensen, J. L. Jensen, "Automatic Target Recognition for Hyperspectral Imagery Using High Order Statistics," IEEE Trans. Aerospace and Electronic Systems, vol. 42, pp. 1372-1385, 2006.
[11]Matteoli, S., M. Diani and G. Corsini, "Improved estimation of local background covariance matrix for anomaly detection in hyperspectral images," Opt. Eng., 49(4), 046201, 2010.
[12]Caefer, C. E., J. Silverman, O. Orthal, D. Antonelli, Y. Sharoni and S. R. Rotman, "Improved covariance matrix for point target detection in hyperspectral images," Opt. Eng., vol. 47(7), 076402, 2008.
[13]Guo, Q., B. Zhang, Q. Ran, L. Gao, J. Li and A. Plaza, "Weighted-RXD and Linear Filter-Based RXD: Improving Background Statistics Estimation for Anomaly Detection," IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 7, pp. 2351-2366, 2014.
[14]Ren, H. and C.-I. Chang, "Automatic Target Recognition in Hyperspectral Imagery," IEEE Trans. Aerospace and Electronic Systems, vol. 39, pp. 1232-1249, 2003.
[15]"USGS Digital Spectral Library," U.S. Geological Survey, a bureau of the U.S. Department of the Interior , [Online]. Available: http://speclab.cr.usgs.gov/spectral-lib.html.
[16]Clark, R. N., G. A. Swayze, A. J. Gallagher, T. V.V. King, and W. M. Calvin, "The U. S. Geological Survey, Digital Spectral Library: Version 1," 1993.
[17]Dell’Endice, F., J. Nieke, B. Koetz, M. E. Schaepman, K. Itten, "Improving radiometry of imaging spectrometers by using programmable spectral regions of interest," ISPRS Journal of Photogrammetry and Remote Sensing, vol. 64, pp. 632-629, 2009.
[18]Yang, R. J., Parallel Computing of Anomaly Detection and Discrimination for Hyperspectral Imagery: a Performance Evaluation, Master Thesis, Dept. Computer Science & Information Engineering, National Central University, Taoyuan City, Taiwan(R. O. C.) 2009.
[19]Strang, G., Introduction to Linear Algebra, Wellesley Cambridge Press, 2009.
[20]Harsanyi, J. C. and C.-I. Chang, "Hyperspectral image classification and dimensionality reduction: an orthogonal subspace projection approach," IEEE Transactions on Geoscience and Remote Sensing, vol. 32, pp. 779-785, 1994.
[21]Chang, C.-I. and D. C. Heinz, "Constrained subpixel target detection for remotely sensed imagery," IEEE Trans. on Geoscience and Remote Sensing, vol. 38, pp. 1144-1159, 2000.
[22]Heinz, D. C. and C.-I. Chang, "Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery," IEEE Trans. on Geoscience and Remote Sensing, vol. 39, pp. 529-545, 2001.
[23]Ren, H., T.H. Wei, and C.H. Hung, "Comparison of weighted least-square approaches for remotely sensed imagery," in Proc. IPPR Conf. on CVGIP 2004, Hwalien, 2004.
[24]Chang, C.-I., Hyperspectral Imaging: Techniques for Spectral Detection and Classification, Springer, 2003.
[25]Ren, H. and C.-I. Chang, "A generalized orthogonal subspace projection approach to unsupervised multispectral image classification," IEEE Trans. on Geosciences and Remote sensing, vol. 38, pp. 2512-2528, 2000.
[26]Muasher, M. J. and D. A. Landgrebe, "The K-L Expansion as an Effective Feature Ordering Technique for Limited Training Sample Size," IEEE Trans. on Geoscience and Remote Sensing, Vols. GE-21, pp. 438-441, 1983.
[27]Jia, X. and J. A. Richards, "Segmented principal components transformation for efficient hyperspectral remote-sensing image display and classification," IEEE Trans. on Geoscience and Remote Sensing, vol. 37, pp. 538-542, 1999.
[28]Chang, C.-I., H. Ren, and S.-S. Chiang, "Real-Time Processing Algorithm for Target Detection and Classification in Hyperspectral Imagery," IEEE Trans. on Geoscience and Remote Sensing, vol. 39, pp. 760-768, 2001.