An ICA Mixture Model Based Approach for Sub-pixel Classification of Hyperspectral Data

Hyperspectral sensors are capable of collecting information in hundreds of contiguous spectral bands to expand the capability of multispectral sensors that use tens of discrete spectral bands. The contiguous bands play a vital role in study of types of vegetation, minerals, forest and soil types. In this paper, an iteratively learning parameter algorithm has been implemented to classify the hyperspectral image in an unsupervised way. The methodology followed in learning the parameters of ICA mixture model (ICAMM) has been discussed and the proportionality constant has been fixed as 0.7 to obtain the linear transformation for estimating the class membership probability of each pixel of a hyperspectral data. The ICAMM algorithm models class distributions as non-Gaussian densities, has been employed for unsupervised classification. Here the data have been transformed into new space in which, the data are as independent as possible by exploiting higher order statistics. This algorithm produces an average overall accuracy of around 65 % and outperforms the conventional K-means clustering and ISODATA.

[1]  Michael S. Lewicki,et al.  Unsupervised image classification, segmentation, and enhancement using ICA mixture models , 2002, IEEE Trans. Image Process..

[2]  Christian F. Beckmann,et al.  Modelling with independent components , 2012, NeuroImage.

[3]  Pramod K. Varshney,et al.  Unsupervised classification of hyperspectral data: an ICA mixture model based approach , 2004 .

[4]  Terrence J. Sejnowski,et al.  Independent Component Analysis Using an Extended Infomax Algorithm for Mixed Subgaussian and Supergaussian Sources , 1999, Neural Computation.

[5]  Pramod K. Varshney,et al.  Decision tree regression for soft classification of remote sensing data , 2005 .

[6]  Russell G. Congalton,et al.  A review of assessing the accuracy of classifications of remotely sensed data , 1991 .

[7]  R. Congalton,et al.  Accuracy assessment: a user's perspective , 1986 .

[8]  David G. Stork,et al.  Pattern Classification (2nd ed.) , 1999 .

[9]  Pramod K. Varshney,et al.  Logistic Regression for Feature Selection and Soft Classification of Remote Sensing Data , 2006, IEEE Geoscience and Remote Sensing Letters.

[10]  Pramod K. Varshney,et al.  Improving Subpixel Classification by Incorporating Prior Information in Linear Mixture Models , 2011, IEEE Transactions on Geoscience and Remote Sensing.

[11]  Jon Atli Benediktsson,et al.  Classification of remote sensing images from urban areas using a fuzzy possibilistic model , 2006, IEEE Geoscience and Remote Sensing Letters.

[12]  Pramod K. Varshney,et al.  ICA mixture model based unsupervised classification of hyperspectral imagery , 2002, Applied Imagery Pattern Recognition Workshop, 2002. Proceedings..

[13]  Mark D. Plumbley.  Algorithms for nonnegative independent component analysis , 2003, IEEE Trans. Neural Networks.

[14]  Erkki Oja,et al.  Independent component analysis: algorithms and applications , 2000, Neural Networks.

[15]  Giles M. Foody,et al.  Incorporating mixed pixels in the training, allocation and testing stages of supervised classifications , 1996, Pattern Recognit. Lett..

[16]  Terrence J. Sejnowski,et al.  ICA Mixture Models for Unsupervised Classification of Non-Gaussian Classes and Automatic Context Switching in Blind Signal Separation , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[17]  Pierre Comon,et al.  Independent component analysis, A new concept? , 1994, Signal Process..

[18]  John A. Richards,et al.  Efficient maximum likelihood classification for imaging spectrometer data sets , 1994, IEEE Trans. Geosci. Remote. Sens..

[19]  Mudassir Rashid,et al.  Multiway independent component analysis mixture model and mutual information based fault detection and diagnosis approach of multiphase batch processes , 2013 .

[20]  John Platt,et al.  Probabilistic Outputs for Support vector Machines and Comparisons to Regularized Likelihood Methods , 1999 .

[21]  Pramod K. Varshney,et al.  A higher order statistical approach to spectral unmixing of remote sensing imagery , 2004, IGARSS 2004. 2004 IEEE International Geoscience and Remote Sensing Symposium.

[22]  G. F. Hughes,et al.  On the mean accuracy of statistical pattern recognizers , 1968, IEEE Trans. Inf. Theory.