投影追蹤方法在高光譜圖像異常點(diǎn)檢測(cè)中的應(yīng)用
Application of Project Pursuit in Hyperspectral Anomaly Detection
-
摘要: 該文提出了一種基于投影追蹤的高光譜圖像異常點(diǎn)檢測(cè)方法。它通過(guò)廣義似然比檢驗(yàn)(GLRT)模型構(gòu)建二元檢測(cè)算子，并利用觀測(cè)數(shù)據(jù)估計(jì)出算子中代表背景的未知參數(shù)，而算子的關(guān)鍵參數(shù)-目標(biāo)參數(shù)是通過(guò)投影追蹤算法搜索異常點(diǎn)得到的。此算法消除了傳統(tǒng)的基于多元統(tǒng)計(jì)模型的目標(biāo)檢測(cè)方法對(duì)先驗(yàn)信息的依賴，增強(qiáng)了算法的實(shí)用性。同時(shí)，投影追蹤方法能有效的提取目標(biāo)參數(shù)，進(jìn)一步提高了異常點(diǎn)檢測(cè)的效果。
-
關(guān)鍵詞:
- 高光譜;投影追蹤;廣義似然比檢驗(yàn);異常檢測(cè)
Abstract: In this paper, a new method of hyperspectral anomaly detection based u* project pursuit is presented. The Generalized Likelihood Ratio Test(GLRT) is used to estab-lish a binary hypotheses detector and estimates the unknown parameters that represent tit* background in the detector from the image. Target information, the key paxameter, is got h using project pursuit approach to search anomaly information. The algorithm reduces the dependence of pre-information, enhances the arithmetic practicability. At the same time, project pursuit approach can extract target information efficiently and improve the effect of anomaly detection. -
Manolakis D, Siracusa C, Marden D, Shaw G. Hyperspectral adaptive matched filter detectors:Practical performance comparison. 2001, SPIE, Vol.4381: 18-33. ［2］Reed I S, Yu Xiaoli. Adaptive multiple-band CFAR detection of an optical pattern with unknown spectral distribution. IEEE Trans. on Acous. Speech and Signal Processing, 1990, ASSP-38(10):1760-1770. ［3］Gillis D, Palmadesso P, Bowles J. An automatic target recognition system for hyperspectral imagery using ORASIS. 2001, SPIE, Vol.4381: 34-43. ［4］Ifarraguerri A, Chang Chein-I. Unsupervised hyperspectral image analysis with project pursuit.IEEE Trans. on Geoscience and Remote Sensing, 2000, GRS-38(6): 2529-2538. ［5］Jimenez L O, Landgrebe D A. Hyperspectral data analysis and supervised feature reduction via projection pursuit. IEEE Trans. on Geoscience and Remote Sensing, 1999, GRS-37(6): 2653-2667. ［6］Chiang Shao-Shan, Chang Chein-I. Unsupervised target detection in hyperspectral image using project pursuit. IEEE Trans. on Geoscience and Remote Sensing, 2001, GRS-39(7): 1380-1391. ［7］江少林,華保盈.利用轉(zhuǎn)動(dòng)慣量導(dǎo)出有限點(diǎn)陣的一種幾何性質(zhì).大學(xué)物理,1998,17(6):20-21. -
計(jì)量
- 文章訪問(wèn)數(shù): 2439
- HTML全文瀏覽量: 120
- PDF下載量: 796
- 被引次數(shù): 0
下載:
