Title :
Spectral-spatial classification for hyperspectral data using SVM and subspace MLR
Author :
Khodadadzadeh, Mahdi ; Jun Li ; Plaza, Antonio ; Ghassemian, Hassan ; Bioucas-Dias, Jose M.
Author_Institution :
Dept. of Technol. of Comput. & Commun., Univ. of Extremadura, Caceres, Spain
Abstract :
This paper presents a new multiple-classifier approach for accurate spectral-spatial classification of hyperspectral images, where the spectral information is exploited by combining probabilistic support vector machines (SVM) and subspace-based multinomial logistic regression (MLRsub) and the spatial information is exploited by means of a Markov random field (MRF) regularizer. The proposed approach is based on the decision fusion of global posterior probability distributions and local probabilities which result from the whole image and the class combinations map respectively. With respect to the SVM or MLRsub algorithms, the proposed method greatly improves the classification accuracy. Our experimental results with real hyperspectral images collected by the NASA Jet Propulsion Laboratory´s Airborne Visible Infra-Red Imaging Spectrometer (AVIRIS) and the Reflective Optics Spectrographic Imaging System (ROSIS), indicate that the proposed multiple-classifier system leads to state-of-the-art classification performance for cases with very limited number of training samples.
Keywords :
Markov processes; decision theory; hyperspectral imaging; image classification; image fusion; random processes; regression analysis; statistical distributions; support vector machines; AVIRIS; MRF regularizer; Markov random field; NASA Jet Propulsion Laboratory; ROSIS; airborne visible infrared imaging spectrometer; decision fusion; global posterior probability distribution; hyperspectral image classification; local probability; multinomial logistic regression; multiple classifier approach; probabilistic SVM; reflective optics spectrographic imaging system; spatial information; spectral information; spectral spatial classification accuracy; subspace-based MLRsub algorithm; support vector machine; Accuracy; Educational institutions; Hyperspectral imaging; Probabilistic logic; Support vector machines; Training; Hyperspectral images; decision fusion; segmentation; spectralspatial classification; support vector machine (SVM);
Conference_Titel :
Geoscience and Remote Sensing Symposium (IGARSS), 2013 IEEE International
Conference_Location :
Melbourne, VIC
Print_ISBN :
978-1-4799-1114-1
DOI :
10.1109/IGARSS.2013.6723247