An Improved Classification Approach Based on Spatial and Spectral Features for Hyperspectral Data
LI Na1, LI Yong-jie1, ZHAO Hui-jie1*, CAO Yang2
1. Key Laboratory of Precision Opto-Mechatronics Technology, Ministry of Education, Beijing University of Aeronautics and Astronautics, Beijing 100191, China 2. Development and Application Center of Command Automation Technology, The 4th Research Institute of CASIC, Beijing 100854, China
Abstract：The spatial correlativity and spectral information are not applied synchronously in the classification model of hyperspectral data. To solve this problem, an improved classification approach based on Markov random field (MRF) theory is proposed in our work. The MRF model based on maximum a posteriori is applied to combine the spectral and spatial information. The probabilistic support vector machine (PSVM) algorithm using pixels spectral information is applied to improve the estimation accuracy of the class conditional probability (CCP) of the spectral energy function, and the efficient belief propagation (EBP) based on multi-accelerated strategy (such as ordinal propagated message strategy, linearized message-updating strategy, and coarse-to-fine approach) is developed in order to solve the problem of the high calculational complexity and time-consumed in the global energy minimum optimization of MRF model. The true hyperspectral data collected by airborne visible infrared imaging spectrometer (AVIRIS) is applied to estimate the performance of the proposed approach in the agricultural demonstration area, Indiana northwest, USA. The performance of the proposed approach is compared with simulated annealing and iterated conditional model. The results illuminate that the average classification accuracy of our method reachs to 95.78%, and the Kappa coefficient is 93.34%, much higher than that of the result by the traditional MRF classification algorithms, and the computational efficiency is improved more than 3 times compared with the belief propagation algorithm.
Key words：Hyperspectral remote sensing;Classification;Markov random field;Probabilistic support vector machine;Efficient belief propagation
李 娜1，李咏洁1，赵慧洁1*，曹 扬2 . 基于光谱与空间特征结合的改进高光谱数据分类算法 [J]. 光谱学与光谱分析, 2014, 34(02): 526-531.
LI Na1, LI Yong-jie1, ZHAO Hui-jie1*, CAO Yang2 . An Improved Classification Approach Based on Spatial and Spectral Features for Hyperspectral Data. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2014, 34(02): 526-531.
[1] YUE Jiang, ZHANG Yi, XU Hai-wei, et al(岳 江，张 毅，徐海威，等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2012, 32(7): 1860. [2] Antonio Plaza, Pablo Martinez, Rosa M Perez，et al. IEEE Transactions on Geoscience and Remote Sensing, 2005, 43(3): 466. [3] Fauvel M, Benediktsson J A, Chanussot J, et al. IEEE Transactions on Geoscience and Remote Sensing, 2008, 46(11): 3804. [4] Bartels M, Wei H. Annual Conference of the Remote Sensing and Photogrammetry Society, UK, 2006. [5] Schweizer S M, Moura J M F. IEEE Trans. on Image Processing, 2001, 10(4): 584. [6] Tso B, Mather P M. New York: Taylor & Francis, 2009. [7] Cao Y, Wei H, Zhao H J. Proc. Pattern Recognition Remote Sensing (PRRS) in ICPR08, USA, 2008: 1. [8] Niu Sidi, Ingle Vinay K, Manolakis Dimitris G, et al. Proc. of SPIE, 2010, 7812: 78120D-1. [9] Szeliski R, Zabih R, Scharstein D, et al. IEEE Transaction on Pattern Analysis and Machine Intelligence, 2008, 30(6): 1068. [10] Gao J, Gunn S, Harris C. Neuro-Computing, 2003, 50: 391. [11] Tarabalka Y, Chanussot J, Benediktsson J A. IEEE Transaction System, Man, and Cybernetics-part B, Cybernetics, 2010, 40: 1267. [12] Platt J C. MIT Press: Advances in Large Margin Classifiers. 1999. [13] Kolmogorov V, Rother C. Proceedings of the European Conference Computer Vision, 2006, 1898: 1. [14] Felzenszwalb P F, Huttenlocher D P. International Journal Computer Vision, 2006, 70: 41. [15] Sun J, Zheng N N, Shum H Y. IEEE Transaction on Pattern Analysis and Machine Intelligence, 2003, 25(7): 787.
