DOI QR코드

DOI QR Code

Two-Dimensional Filtering Through the Radon Transform

라돈변환을 이용한 2차원 필터링

  • 원중선 (연세대학교 지구시스템과학과)
  • Published : 1998.06.01

Abstract

The Radon transform has been widely used in various techniques of digital image processing such as the computerized topography, lineament analysis in a remotely sensed image, slant-stack processing of seismic data, and so on. Compared to the Fourier transform, the utility of two-dimensional convolutional or correlational properties of the Radon transform, however, has been underestimated. We show that the two-dimensional convolution and correlation is respectively reduced to be one-dimensional convolution and correlation with respect to ρ in the Radon space. Therefore, one can achieve a two dimensional filtering by applying a simple one-dimensional convolution in the Radon space followed by an inverse Radon transform. Tests of the approach using FIR filters are carried out specifically for enhancing the ship wake in a RADARSAT SAR image. The test results demonstrate that the two-dimensional filtering through the Radon transform effectively enhance the ship wake features as well as reducing sea speckle in the image. Although two-dimensional convolution and correlation through the Radon transform are not so much useful as those through the courier transform in views of efficiency and effectiveness, it can be utilized to improve the quality of a digitally processed output when the process should be accompanied by the Radon transform such as topography and lineament analysis of SAR image.

라돈변환은 디지털 토모그라피. 원격탐사 영상의 선구조 분석, 탄성파탐사 자료의 경사중첩 등 다양한 분야에서의 디지털 영상자료 처리에 폭넓게 이용되어 왔다. 그러나 라돈변환의 2차원 콘볼류션이나 상관관계 계산은 푸리에변환과 비교해 볼 때 그 효용성에 대해서는 거의 논의되지 못하였다. 본 논문에서는 2차원 콘볼류션이나 상관관계 계산은 라돈공간에서 ρ에 대한 1차원 콘볼류션이나 상관관계로 타원이 낮아진다는 것을 유도하였다. 이를 이용하면 영상공간에서의 2차원 필터링을 라돈공간에서의 단순한 1타원 필터링으로 효과적으로 수행할 수 있다. 본 논문에서는 FIR 필터를 이용한 라돈공간에 서의 1차원 필터링을 RADARSAT SAR 영상에서 선적에 의한 구조를 증강시키데 적용하였다. 적용결과 라돈변환을 이용한 2타원 필터링은 스펙클 잡음을 줄일 뿐만 아니라 효과적으로 선적에 의한 구조를 증강시키는 것으로 나타났다. 비록 라돈변환을 이용하는 2차원 필터링은 푸리에변환을 이용하는 방법에 비하여 효율성과 효과도가 떨어져 단순한 필터링 목적으로는 적합치 못하나, 토모그라피나 SAR 영상을 이용하여 선구조 분석을 실시하는 경우와 같이 라돈변환이 필수적으로 적용되는 경우 라돈공간에서의 2차원 필터링을 부수적으로 적용하면 효과적으로 활용될 수 있다.

Keywords

References

  1. IEEE Trans. Acoust., Speech, and Signal Processing v.35 no.2 Discrete Radon transform Beylkin, G.
  2. IEEE Trans. Geosci. Remote Sensing v.33 no.1 Localized Radon transform-based detection of ship wakes in SAR images Copeland, A.C;G.R. Ravichandran;M.M. Trivedi
  3. The Radon Transform and Some of Its Applications Deans, S.R.
  4. IEEE Trans. Image Processing v.4 no.1 Multiresolution tomographic reconstruction using wavelets Delaney, A.H.;Y. Bresler
  5. Comm. ACM v.15 no.1 Use of the Hough transformation to detect lines and curves in pictures Duda, R.O.;P.E. Hart
  6. Multidimensional Digital Signal Processing Dudgeon, D. E.;R. M. Rersereau
  7. Geophysics v.49 no.8 The Radon transform and its properties Durrani, T.S.;D. Bisset
  8. Proc. IGARSS88 Automated ship and ship wake detection in spaceborne SAR images from coastal regions Eldhuset, K.
  9. Proc. ISPRSXVⅢ Congress The EPOS filter: A comparision with some well-known speckle reduction techniques Hagg, W.;M. Sties
  10. Computer Vision, Graphics, and Image Processing v.44 A survey of the Hough transform Illingworth, J.;J. Kittler
  11. Digital Image Processing Techniques Image Reconstruction from Projection Kak, A.C.;M.D. Ekstrome(ed.)
  12. Trans. AGU v.7 no.3 Mystery ship detected in SAR image, EOS Liu, A.K.;C.Y. Peng;Y.-S. Chang
  13. Pattern Recognition Letters v.4 Linear feature detection and enhancement in noisy images via the Radon transform Murphy, L.M.
  14. Berichte Sachsische Akademie der Wissenschaften v.69 Uber die Bestimmung von Funktionen durch ihre Integralwerte langs gewisser Manningfaltigkeiten Radon, J.
  15. IEEE Trans. Geosci. Remote Sensing v.28 no.4 Applicaiton of Radon transform techniques to wake detection in Seasat-A SAR images Rey, M.T.;J.K. Tunaley;J.T. Folinsbee;P.A. Jahans;J.A. Dixon;M.R. Vant
  16. Digital Picture Processing(2nd ed.) v.1 Rosenfeld, A.;A.C. Kak
  17. Geophysics: The Leading Edge of Exploration Noise elimination and the Radon transform Russell, B.;D. Hampson;J. Chun
  18. Geophysics: The Leading Edge of Exploratin Noise elimination and the Radon transform Russell, B.;D. Hampson;J. Chun
  19. IEEE Trans. Signal Processing v.41 no.12 Image reconstruction from projections under wavelet constraints Sahiner, B.;A.E. Yagle
  20. Proc. IGARSS89 v.Ⅰ Edge following as graph searching and Hough transform algorithm for lineament detection Wang, J.;P.J. Howarth
  21. J. Korean Soc. Remote Sensing v.13 no.3 Speckle reduction of multiple beam mode RADARSAT SAR data using wavelet transform Won, J.S.;J.H. Ryu;S.W. Kim