Journal of the Institute of Electronics Engineers of Korea SC (전자공학회논문지SC)

The Institute of Electronics and Information Engineers (대한전자공학회)
권호 표지

Hyperspectral Image Fusion for Tumor Detection

초분광 영상 융합을 이용한 종양인식

  • Xu Cheng-Zhe (Department of Communication Engineering Myongji University) ;
  • Kim In-Taek (Department of Communication Engineering Myongji University)
  • Published : 2006.07.01
Download PDF
⟨ Previous Next ⟩

Abstract

This paper presents a method for detecting tumors on chicken carcasses by fusion of hyperspectral fluorescence and reflectance images. Classification of normal skin and tumor is performed by the image obtain 어 from optimal band ratio which minimizes the overlapping area of PDFs for normal skin and tumor. This method yields four feature images, each of them represents the ratio of two intensity values from a pixel. Classification is achieved by applying ISODATA to each pixel from the feature images. For the analysis of reflectance image, band selection method is proposed based on the information quantity, many effective features are acquired for the classification by defining the linear transformation selecting the projection axis, accordingly, accurate interpretation of images is possible in the reflectance image and automatic feature selection method is realized. Feature images from reflectance images are also classified by ISODATA and combined with the result from fluorescence images. Experimental result indicates that improved performance in term of reducing false detection rate is observed.

본 논문에서는 초분광 형광영상과 반사영상 융합을 이용한 닭의 종양인식방법을 제안하였다. 형광영상에 밴드비율을 적용하여 피부의 정상과 종양부분을 구분한다. 이를 위해 각각 부분의 확률밀도함수의 중첩된 면적을 최소화하는 방법을 사용하였다. 이 방법으로 획득한 4개의 특정영상에 분할-합병법을 적용하여 형광영상 분류결과를 얻었다. 반사영상 분석에서는 단일 밴드가 정보량에 주는 영향에 근거하여 밴드 선택 방법을 제안하였다. 학습데이터에 의해 투영 축을 선택하는 선형변환을 정의함으로써 영상분류에 효과적인 많은 특징을 확보하였다. 이에 따라 반사영상에서도 세밀한 영상의 해석이 가능하였고 특징 선택의 자동화를 실현하였다. 반사영상에서 획득한 특정영상도 분할-합병법으로 분류하였으며 형광영상의 분류결과와 융합하여 종양을 인식하였다. 모의실험을 통해 제안한 방법은 기존의 방법에 비해 오인식이 낮음을 확인하였다.

Keywords

References

  1. Zhou, L., Kambhamettu, C., Goldgof, D.B., Fluid structure and motion analysis from muti-spectrum 2D clould image sequences, Computer Vision and Pattern Recognition, 2000. Proceedings. IEEE Conference on, Vol. 2, pp. 744-751, 2000 https://doi.org/10.1109/CVPR.2000.854949
  2. Chen, Y. R., Classifying diseased poultry carcasses by visible and near-IR reflectance spectroscopy, Optics in Agriculture and forestry, SPIE 1836, pp. 46-55, 1993 https://doi.org/10.1117/12.144043
  3. Chao, K., Y. R. Chen, W. R. Hrushka, and F. B. Gwozdz.,On-line inspection of poultry carcasses by a dual-camera syste, J. Food. Eng. 51, pp. 185-192, 2002 https://doi.org/10.1016/S0260-8774(01)00051-6
  4. Huebschman, M. L., Schultz, R.A., Garner, H. R., Characteristics and capabilities of the hyperspectral imaging microscope, IEEE Engineering in Medicine and Biology Magazine, vol. 21 Issue 4, pp. 104-117, July-Aug, 2002 https://doi.org/10.1109/MEMB.2002.1032647
  5. Kim, I., M.S. Kim, Y.R. Chen, and S.G. Kong, Detection of skin tumors on chicken carcasses using hyperspectral fluorescence imaging, Transactions of ASAE, Vol. 47(5), pp. 1785-1792, 2004 https://doi.org/10.13031/2013.17595
  6. 김한열, 김인택, 하이퍼스펙트럴 영상 분석, 대한전기학회 논문지 Vol. 52D, No. 11, pp. 634-643, Nov. 2003
  7. Bowman, A. W. and A. Azzalini, Applied Smoothing Techniques for Data Analysis, Oxford University Press, 1997
  8. Menahem Friedman, Abraham Kandel, Introduction to Pattern Recognition, Imperial college Press, pp. 81-85, 1999
  9. Tsuhan Chen, Hsu, Y.J., Xiaoming Liu, Wende Zhang, Principle component analysis and its variants for biometrics, Image Processing, 2002. Proceedings. 2002 International Conference on, Vol. 1, pp. 61-64, 2002 https://doi.org/10.1109/ICIP.2002.1037959
  10. Goodman, S., Hunter, A., Feature extraction algorithms for pattern classification, Artificial Neural Networks, 1999. ICANN 99. Ninth International Conference on (conf. Publ. No. 470), Vol. 2, pp. 738-742, 1999 https://doi.org/10.1049/cp:19991199