神經(jīng)網(wǎng)絡(luò)敏感性分析的高光譜遙感影像降維與分類方法
doi: 10.11999/JEIT160052
中央高?;究蒲袠I(yè)務(wù)費項目(2014B13214, 2015B 26914),十二五國家科技支撐計劃項目(2015BAB07B03),河海大學(xué)國家級大學(xué)生創(chuàng)新訓(xùn)練計劃項目(201610294061)
Dimension Reduction and Classification of Hyperspectral Remote Sensing Images Based on Sensitivity Analysis of Artificial Neural Network
The Fundamental Research Funds for the Central Universities (2014B13214, 2015B26914), The Projects in the National Science Technology Pillar Program during the Twelfth Five-year Plan Period (2015BAB07B03), The National Undergraduate Training Program for Innovation and Entrepreneurship of Hohai University (201610294061)
-
摘要: 高光譜遙感影像由于其巨大的波段數(shù)直接導(dǎo)致信息的高冗余和數(shù)據(jù)處理的復(fù)雜,這不僅帶來龐大的計算量,而且會損害分類精度。因此,在對高光譜影像進行處理、分析之前進行降維變得非常必要。神經(jīng)網(wǎng)絡(luò)敏感性分析可以用于對模型的簡化降維,該文將該方法運用于高光譜遙感影像降維中,通過子空間劃分弱化波段之間的相關(guān)性,利用差分進化算法(DE)優(yōu)化神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu),采用Ruck敏感性分析方法剔除掉對分類貢獻較小的波段,從而實現(xiàn)降維。最后,采用AVIRIS影像進行實驗,所提算法相比其他相近的降維與分類方法能獲得更高的分類精度,達到85.83%,比其他相近方法中最優(yōu)方法高出0.31%。
-
關(guān)鍵詞:
- 神經(jīng)網(wǎng)絡(luò)敏感性分析 /
- 高光譜遙感影像降維 /
- 子空間劃分 /
- 差分進化 /
- Ruck敏感性分析
Abstract: The high dimensions of hyperspectral remote sensing images will cause the redundancy of information and complexity of data processing, which also brings tremendous computing workload and damages application accuracy. Therefore, before the analysis of hyperspectral image processing, it is necessary to reduce the high dimensions of hyperspectral data. The Sensitivity Analysis (SA) of artificial neural network can be used in dimension reduction of the model. Now the Sensitivity Analysis of artificial neural network is applied to dimension reduction for hyperspectral remote sensing images in the paper. First of all, all bands are divided into several groups as long as a lower correlation exists between adjacent bands. Furthermore, Differential Evolution (DE) algorithm is used for optimizing neural network structure. Moreover, the bands which make small contribution will be given up based on Ruck sensitivity analysis method. Finally, experiments are conducted with AVIRIS images. The results show that the proposed method can get high classification accuracy of 85.83% at small training samples, 0.31% higher than the best one among other similar methods of dimension reduction and classification. -
杜培軍, 譚琨, 夏俊士. 高光譜遙感影像分類與支持向量機應(yīng)用研究[M]. 北京: 科學(xué)出版社, 2012: 6-35. DU Peijun, TAN Kun, and XIA Junshi. Classification of Hyperspectral Remote Sensing Images and Applied Research of SVM[M]. Beijing: Science Press, 2012: 6-35. 童慶禧, 張兵, 鄭蘭芬. 高光譜遙感原理、技術(shù)及應(yīng)用[M]. 北京: 高等教育出版社, 2006: 33-56. TONG Qingxi, ZHANG Bing, and ZHENG Lanfen. Hyperspectral Remote Sensing-Principles, Techniques and Applications[M]. Beijing: Higher Education Press, 2006: 33-56. 吳倩, 張榮, 徐大衛(wèi). 基于稀疏表示的高光譜數(shù)據(jù)壓縮算法[J]. 電子與信息學(xué)報, 2015, 37(1): 78-84. doi: 10.11999/ JEIT140214. WU Qian, ZHANG Rong, and XU Dawei. Hyperspectral data compression based on sparse representation[J]. Journal of Electronics Information Technology, 2015, 37(1): 78-84. doi: 10.11999/JEIT140214. GAO Hongmin, XU Lizhong, LI Chenming, et al. A new feature selection method for hyperspectral image classification based on simulated annealing genetic algorithm and choquet fuzzy integral[J]. Mathematical Problems in Engineering, 2013: 1-14. doi: 10.1155/2013/537268. GAO Lianru, LI Jun, KHODADADZADEH M, et al. Subspace-based support vector machines for hyperspectral image classification[J]. IEEE Geoscience and Remote Sensing Letters, 2015, 12(2): 349-353. doi: 10.1109/LGRS.2014. 2341044. GURRAM P and KWON H. Coalition game theory based feature subset selection for hyperspectral image classification [C]. IEEE International Geoscience and Remote Sensing Symposium, Quebec, Canada, 2014: 3446-3449. FALCO N, BENEDIKTSSON J A, and BRUZZONE L. A study on the effectiveness of different independent component analysis algorithms for hyperspectral image classification[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014, 7(6): 2183-2199. doi: 10.1109/JSTARS.2014.2329792. 姜宇, 肖鴻, 劉興鵬, 等. BP神經(jīng)網(wǎng)絡(luò)在異向介質(zhì)基本結(jié)構(gòu)分析中的應(yīng)用[J]. 電子與信息學(xué)報, 2010, 32(1): 195-198. doi: 10.3724/SP.J.1146.2008.01703. JIANG Yu, XIAO Hong, LIU Xingpeng, et al. Applications of BP neural network in analyzing metamaterials elemental basic structure[J]. Journal of Electronics Information Technology, 2010, 32(1): 195-198. doi: 10.3724/SP.J.1146. 2008.01703. 張兵, 高連如. 高光譜圖像分類與目標(biāo)探測[M]. 北京: 科學(xué)出版社, 2011: 85-101. ZHANG Bing, GAO Lianru. Hyperspectral Image Classification and Target Detection[M]. Beijing: Science Press, 2011: 85-101. 蔡毅, 邢巖, 胡丹. 敏感性分析綜述[J]. 北京師范大學(xué)學(xué)報(自然科學(xué)版), 2008, 44(1): 9-16. CAI Yi, XING Yan, and HU Dan. On sensitivity analysis[J]. Journal of Beijing Normal University(Natural Science), 2008, 44(1): 9-16. 張軍, 劉祖強, 張正祿, 等. 基于神經(jīng)網(wǎng)絡(luò)和模糊評判的滑坡敏感性分析[J]. 測繪科學(xué), 2012, 37(3): 59-62. ZHANG Jun, LIU Zuqiang, ZHANG Zhenglu, et al. Susceptibility of landslide based on artificial neural networks and fuzzy evaluating model[J]. Science of Surveying and Mapping, 2012, 37(3): 59-62. ZHANG Junping, ZHANG Ye, ZOU Bin, et al. Fusion classification of hyperspectral image based on adaptive subspace decomposition[C]. IEEE International Conference on Image Processing, Vancouver, BC, Canada, 2000, 3: 472-475. YU Feng and XU Xiaozhong. A short-term load forecasting model of natural gas based on optimized genetic algorithm and improved BP neural network[J]. Applied Energy, 2014, 134: 102-113. doi: 10.1016/j.apenergy.2014.07.104. LIU Ruixin, ZHANG Xiaodong, ZHANG Lu, et al. Bitterness intensity prediction of berberine hydrochloride using an electronic tongue and a GA-BP neural network[J]. Experimental and Therapeutic Medicine, 2014, 7(6): 1696-1702. doi: 10.3892/etm.2014.1614. 錢文江, 李同春, 丁林. 基于改進BP神經(jīng)網(wǎng)絡(luò)的庫區(qū)滲漏量敏感性分析[J]. 三峽大學(xué)學(xué)報(自然科學(xué)版), 2012, 34(6): 23-27. QIAN Wenjiang, LI Tongchun, and DING Lin. Sensitivity analysis of reservoirs seepage discharge based on improved BP network[J]. Journal of China Three Gorges University (Natural Science), 2012, 34(6): 23-27. WANG Lin, ZENG Yi, and CHEN Tao. Back propagation neural network with adaptive differential evolution algorithm for time series forecasting[J]. Expert Systems with Applications, 2014, 42(2): 855-863. doi: 10.1016/j.eswa.2014. 08.018. RUCK D W, ROGERS S K, and KABRISKY M. Feature selection using a multilayer perceptrons[J]. Journal of Neural Network Computing, 1990, 2(2): 40-48. ZURADA J M, MALINOWSKI A, and USUI S. Perturbation method for deleting redundant inputs of perceptron networks[J]. Neurocomputing, 1997, 14(2): 177-193. doi: 10.1007/978-3-662-45652-1_35. -
計量
- 文章訪問數(shù): 1697
- HTML全文瀏覽量: 184
- PDF下載量: 603
- 被引次數(shù): 0