Band selection algorithm for orb feature detector in hyperspectral earth remote sensing data

G.R. Kasoev, G.A. Scherbina

Moscow Institute of Physics and Technology,

RF, Moscow, Kerchenskaya St., 28A

DOI: 10.33075/2220-5861-2024-2-95-103

UDC 004.93’11                                                          



ORB feature detector is used for stitching of images from different optical channels of two-slit hyperspectrometer. It is essential to choose a subset of all available spectral bands to avoid computational complexity. Two band selection algorithms are discussed in the scope of this work. Algorithms are spectral cluster center search (ECA) and estimation of information sufficiency and redundancy based on pairwise correlation (BCA). Algorithms are modified with spectral distance between bands to ensure search for spectrally robust features. Testing with EO-1 Hyperion data is conducted. Results show that modification increases the number of detected feature and accuracy of matched pairs by 10-20% for ECA algorithm without substantially increasing computational time.

Keywords: hyperspectrometer, feature, ORB, image stitching, ECA, BCA.

Full text in PDF(RUS)


  1. Hennessy A., Clarke K., and Lewis M. Hyperspectral Classification of Plants: A Review of Waveband Selection Generalisability. Remote Sens, 2020, Vol. 12, No. 1, 113 p.
  2. Pearlman J., Barry P., Segal C., Shepanski J., Beiso D., and Carman S. Hyperion, a Space-Based Imaging Spectrometer. Geoscience and Remote Sensing, IEEE Transactions, 2003, Vol. 41, pp. 1160–1173.
  3. Belyaev M.Yu., Korotkov D.M., Kuzmichev A.S., Nikolenko A.A., Cheremisin M.V., Shibanov S.Yu., Shcherbakov M.V., and Shcherbina G.A. Distancionnoe zondirovanie Zemli s rossijskogo segmenta MKS s ispol’zovaniem perspektivnoj nauchnoj apparatury giperspektrometr (Earth remote sensing from Russian segment of ISS with perspective science apparatus hyperspectrometer). XVII Vserossijskaya konferenciya “Sovremennye problemy distancionnogo zondirovaniya Zemli iz kosmosa” (XVII All-Russia Conference “Modern problems of remote sensing of the Earth from space”.). Moscow, Russia, 11–15 November 2019, Book of Abstracts, p. 508.
  4. Rublee E., Rabaud V., Konolige K., and Bradski G. ORB: An efficient alternative to SIFT or SURF. 2011 International Conference on Computer Vision, Barcelona, Spain, 2011, pp. 2564–2571.
  5. Wang R., Zhang W., Shi Y., Wang X., and Cao W. GA-ORB: A new efficient feature extraction algorithm for multispectral images based on geometric algebra. IEEE access, 2019, Vol. 7, pp. 71235–71244.
  6. Sun W. and Du Q. Hyperspectral Band Selection: A Review. IEEE Geoscience and Remote Sensing Magazine, 2019, Vol. 7, No. 2, pp. 118–139.
  7. Du Q. and Yang H. Similarity-Based Unsupervised Band Selection for Hyperspectral Image Analysis. IEEE Geoscience and Remote Sensing Letters, 2008, Vol. 5, No. 4. pp. 564–568.
  8. Dos Santos L. C. B., Guimarães S. J. F., and dos Santos J. A. Efficient Unsupervised Band Selection Through Spectral Rhythms. IEEE Journal of Selected Topics in Signal Processing, 2015, Vol. 9, No. 6, pp. 1016–1025.
  9. Wang L., Jia X., and Zhang Y. A novel geometry-based feature-selection technique for hyperspectral imagery. IEEE Geoscience and Remote Sensing Letters, 2007, Vol. 4, No. 1, pp. 171–175.
  10. Zhang L., Zhong Y., Huang B., Gong J., and Li P. Dimensionality reduction based on clonal selection for hyperspectral imagery. IEEE Transactions on Geoscience and Remote Sensing, 2007, Vol. 45, No. 12, pp. 4172–4186.
  11. Shi A., Gao H., He Z., Li M., and Xu L. A hyperspectral band selection based on game theory and differential evolution algorithm. International Journal on Smart Sensing and Intelligent Systems, 2016, Vol. 9, No. 4, pp. 1971–1990.
  12. Su H., Du Q., Chen G., and Du P. Optimized hyperspectral band selection using particle swarm optimization. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014, Vol. 7, No. 6, pp. 2659–2670.
  13. Sun K., Geng X., and Ji L. A band selection approach for small target detection based on CEM. International Journal of Remote Sensing, 2014, Vol. 35, No. 13, pp. 4589–4600.
  14. Sun K., Geng X., Ji L., and Lu Yun-teng. A new band selection method for hyperspectral image based on data quality. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014. Vol. 7, No. 6, pp. 2697–2703.
  15. Geng X., Sun K., Ji L., Tang H., and Zhao Y. Joint skewness and its application in unsupervised band selection for small target detection. Scientific reports. 2015. Vol. 5, No. 1, p. 9915.
  16. Kim J. H., Kim J., Yang Y., Kim S., and Kim H. S. Covariance-based band selection and its application to near-real-time hyperspectral target detection. Optical Engineering, 2017, Vol. 56, No. 5, pp. 053101-053101.
  17. Bajcsy P. and Groves P. Methodology for hyperspectral band selection. Photogrammetric Engineering & Remote Sensing, 2004, Vol. 70, No. 7, pp. 793–802.
  18. Chang C. I., Du Q., Sun T. L., and Althouse M. A joint band prioritization and band-decorrelation approach to band selection for hyperspectral image classification. IEEE transactions on geoscience and remote sensing, 1999, Vol. 37, No. 6, pp. 2631–2641.
  19. Chang C. I. and Wang S. Constrained band selection for hyperspectral imagery. IEEE transactions on geoscience and remote sensing, 2006, Vol. 44, No. 6, pp. 1575–1585.
  20. Li H. C., Chang C.I., Wang L., and Li Y. Constrained multiple band selection for hyperspectral imagery. 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS). IEEE, 2016, pp. 6149-6152.
  21. Zhang W., Li X., and Zhao L. A fast hyperspectral feature selection method based on band correlation analysis. IEEE Geoscience and Remote Sensing Letters, 2018, Vol. 15, No. 11, pp. 1750–1754.
  22. Sun K., Geng X., and Ji L. Exemplar component analysis: A fast band selection method for hyperspectral imagery. IEEE Geoscience and Remote Sensing Letters, 2014, Vol. 12, No. 5, pp. 99–1002.