Unsupervised band removal leading to improved classification accuracy of hyperspectral images

ABSTRACT

Remotely-sensed images of the earth's surface are used across a wide range of industries and applications including agriculture, mining, defence, geography and geology, to name but a few. Hyperspectral sensors produce these images by providing reflectance data from the earth's surface over a broad range of wavelengths or bands. Some of the bands suffer from a low signal-to noise ratio (SNR) and do not contribute to the subsequent classification of pixels within the hyperspectral image. Users of hyperspectral images typically become familiar with individual images or sensors and often manually omit these bands before classification.We propose a process that automatically determines the spectral bands that may not contribute to classification and removes these bands from the image. Removal of these bands improves the classification performance of a well-researched hyperspectral test image by over 10% whilst reducing the size of the image from a data storage perspective by almost 30%. The process does not rely on prior knowledge of the sensor, the image or the phenomenology causing the SNR problem.In future work, we aim to develop compression algorithms that incorporate this process to achieve satisfactory compression ratios whilst maintaining acceptable classification accuracies.

REFERENCES

• 1: Shah, C. A., Watanachaturaporn, P., Arora, M. K. and Varshney, P. K. (2003): Some Recent Results on Hyperspectral Image Classification, IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, Greenbelt, MD, USA.
• 2: Tsai, F. and Philpot, W.D. (2002): A Derivative-Aided Hyperspectral Image Analysis System for Land-Cover Classification, IEEE Transactions on Geoscience and Remote Sensing, 40:2:416-425.
• 3: AVIRIS - General Overview, NASA http://aviris.jpl.nasa.gov/. Accessed 27 Jul 2005.
• 4: Chintan A. Shah , Manoj K. Arora , Stefan A. Robila , Pramod K. Varshney, ICA Mixture Model based Unsupervised Classification of Hyperspectral Imagery, Proceedings of the 31st Applied Image Pattern Recognition Workshop on From Color to Hyperspectral: Advancements in Spectral Imagery Exploitation, p.29-35, October 16-18, 2002
• 5: 92AV3C: Hyperspectral Test Image Data Pack, Purdue University ftp://ftp.ecn.purdue.edu/biehl/MultiSpec/92AV3C. Accessed 31 July 2005.
• 6: Faulconbridge, R.I., Pickering, M.R., Ryan, M.J. and Jia, X. (2003): A New Approach to Controlling Compression-Induced Distortion of Hyperspectral Images, Proceedings of The International Geoscience and Remote Sensing Society Conference, Toulouse, France, III:1830-1832.
• 7: Anderberg, M.R. (1973), Cluster Analysis for Applications, Academic Press, New York.
• 8: Hardy, A. and Lallemand, P. (2002): Determination of the Number of Clusters for Symbolic Objects Described by Interval Variables. In Classification, Clustering and Data Analysis, 311-316, Springer-Verlag, Berlin.
• 9: Sweet, J.N. (2003): The spectral similarity scale and its application to the classification of hyperspectral remote sensing data, IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 92-99, Greenbelt, MD, USA.
• 10: J. A. Richards, Remote Sensing: Digital Image Analysis, Springer-Verlag New York, Inc., Secaucus, NJ, 1986
• 11: Purdue/LARS Multispec©, Remote Sensing Freeware Software, Purdue Research Foundation, http://dynamo.ecn.purdue.edu/~biehl/Multispec/. Accessed 27 Jul 2005.
• 12: Landgrebe, D. and Biehl, L. (2001): An Introduction to Multispec, p133, Purdue Research Foundation, Indiana, USA.

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.5 PATTERN RECOGNITION
      I.5.2 Design Methodology
          Subjects: Classifier design and evaluation

Additional Classification:
  I. Computing Methodologies
  I.4 IMAGE PROCESSING AND COMPUTER VISION
        I.4.10 Image Representation
      I.4.2 Compression (Coding)

General Terms:
Algorithms, Design, Performance

Keywords:
SNR, classification, hyperspectral