DOI QR코드

DOI QR Code

초분광 표적 탐지를 위한 L2,1-norm Regression 기반 밴드 선택 기법

Band Selection Using L2,1-norm Regression for Hyperspectral Target Detection

  • 김주창 (한국과학기술원 전기 및 전자공학부) ;
  • 양유경 (국방과학연구소) ;
  • 김준형 (국방과학연구소) ;
  • 김준모 (한국과학기술원 전기 및 전자공학부)
  • 투고 : 2017.02.22
  • 심사 : 2017.08.07
  • 발행 : 2017.10.30

초록

초분광 영상을 이용한 표적 탐지를 수행할 때에는 인접한 분광 밴드의 중복성의 문제 및 고차원 데이터로 인해 발생하는 방대한 계산량의 문제점을 해결하기 위한 특징 추출 과정이 필수적이다. 본 연구는 기계 학습 분야의 특징 선택 기법을 초분광 밴드 선택에 적용하기 위해 $L_{2,1}$-norm regression 모델을 이용한 새로운 밴드 선택 기법을 제안하였으며, 제안한 밴드 선택 기법의 성능 분석을 위해 표적이 존재하는 초분광영상을 직접 촬영하고 이를 바탕으로 표적 탐지를 수행한 결과를 분석하였다. 350 nm~2500 nm 파장 대역에서 밴드 수를 164개에서 약 30~40개로 감소시켰을 때 Adaptive Cosine Estimator(ACE) 탐지 성능이 유지되거나 향상되는 결과를 보였다. 실험 결과를 통해 제안한 밴드 선택 기법이 초분광 영상에서 탐지에 효율적인 밴드를 추출해 내며, 이를 통해 성능의 감소 없이 데이터의 차원 감소를 수행할 수 있어 향후 실시간 표적 탐지 시스템의 처리 속도 향상에 도움을 줄 수 있을 것으로 보인다.

When performing target detection using hyperspectral imagery, a feature extraction process is necessary to solve the problem of redundancy of adjacent spectral bands and the problem of a large amount of calculation due to high dimensional data. This study proposes a new band selection method using the $L_{2,1}$-norm regression model to apply the feature selection technique in the machine learning field to the hyperspectral band selection. In order to analyze the performance of the proposed band selection technique, we collected the hyperspectral imagery and these were used to analyze the performance of target detection with band selection. The Adaptive Cosine Estimator (ACE) detection performance is maintained or improved when the number of bands is reduced from 164 to about 30 to 40 bands in the 350 nm to 2500 nm wavelength band. Experimental results show that the proposed band selection technique extracts bands that are effective for detection in hyperspectral images and can reduce the size of the data without reducing the performance, which can help improve the processing speed of real-time target detection system in the future.

키워드

참고문헌

  1. Du, Q. and H. Yang, 2008. Similarity-based unsupervised band selection for hyperspectral image analysis, IEEE Geoscience and Remote Sensing Letters, 5(4): 564-568. https://doi.org/10.1109/LGRS.2008.2000619
  2. Gerg, I., 2010. An evaluation of three endmember extraction algorithms: ATGP, ICA-EEA & VCA, Proc. of 2010 2nd Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing, Reykjavik, Iceland, Jun.14-16, pp. 1-4.
  3. Harsanyi, J. C., 1993. Detection and classification of subpixel spectral signatures in hyperspectral image sequences, Diss. University of Maryland Baltimore County, 6372
  4. Kraut, S., L. L. Scharf, and R. W. Butler, 2005. The adaptive coherence estimator: A uniformly most-powerful-invariant adaptive detection statistic, IEEE Transactions on Signal Processing, 53(2): 427-438. https://doi.org/10.1109/TSP.2004.840823
  5. Manolakis, D., D. Marden, and G. A. Shaw, 2003. Hyperspectral image processing for automatic target detection applications, Lincoln Laboratory Journal, 14(1): 79-116.
  6. Nie, F., H. Huang, X. Cai, and C. H. Ding, 2010. Efficient and robust feature selection via joint $L_{2,1}$-norms minimization, In Advances in neural information processing systems, 1813-1821.
  7. Ren, H. and C. Chang, 2003. Automatic spectral target recognition in hyperspectral imagery, IEEE Transactions on Aerospace and Electronic Systems, 39(4): 1232-1249. https://doi.org/10.1109/TAES.2003.1261124
  8. Shaw, G. A. and H. K. Burke, 2003. Spectral imaging for remote sensing, Lincoln Laboratory Journal, 14(1): 3-28.
  9. Sun, K., X. Geng, and L. Ji, 2015. A new sparsitybased band selection method for target detection of hyperspectral image, IEEE Geoscience and Remote Sensing Letters, 12(2): 329-333. https://doi.org/10.1109/LGRS.2014.2337957
  10. Tibshirani, R., 1996. Regression shrinkage and selection via the lasso, Journal of the Royal Statistical Society. Series B (Methodological), 267-288.
  11. Keshava, N., 2004. Distance metrics and band selection in hyperspectral processing with applications to material identification and spectral libraries, IEEE Transactions on Geoscience and Remote Sensing, 42(7): 1552-1565. https://doi.org/10.1109/TGRS.2004.830549
  12. Winkelmann, M., 2015. Spectral analysis of the vegetative background in the NIR and SWIR spectral range, Proc. of SPIE Security+ Defence. International Society for Optics and Photonics, Toulouse, France, Sep. 11-14, vol. 9653, pp. 96530D-96530D.
  13. Yang, H., Q. Du, H. Su, and Y. Sheng, 2011. An efficient method for supervised hyperspectral band selection, IEEE Geoscience and Remote Sensing Letters, 8(1): 138-142. https://doi.org/10.1109/LGRS.2010.2053516
  14. Zhang, J., Y. Cao, L. Zhuo, C. Wang, and Q. Zhou, 2015. Improved band similarity-based hyperspectral imagery band selection for target detection, Journal of Applied Remote Sensing, 9(1): 095091-095091. https://doi.org/10.1117/1.JRS.9.095091