KSCE Journal of Civil and Environmental Engineering Research (대한토목학회논문집)

Korean Society of Civil Engeneers (대한토목학회)
권호 표지

A Rigorous Geometric Rectification of RADARSAT SAR Imagery Using a Single Control Point

단일 기준점을 이용한 RADARSAT SAR 영상의 엄밀 기하보정

Sohn, Hong-Gyoo;Song, Yeong-Sun;Kim, Gi-Hong;Bang, Soo-Nam
손홍규;송영선;김기홍;방수남

  • Published : 2004.01.31

⟨ Previous Next ⟩

Abstract

For accurate geometric modeling of SAR imagery the relationship between image coordinates and ground coordinates must be precisely established. The most tedious and time-consuming step in geocoding of SAR imagery is the precise identification of ground control points from SAR imagery. This task is a lot more difficult than that of optical sensor imagery due to the existence of speckle noise in SAR imagery. Moreover since SAR imagery is frequently used for the analysis of natural disaster such as flood area analysis, earthquake monitoring, and forest fire monitoring due to its day and night and all-weather capabilities, the real time geometric modeling of SAR imagery is essential for the rapid analysis. In this paper the rigorous geometric modeling of SAR imagery using a single ground control point is performed to reduce processing time. For achieving the goal we utilized a single ground control point together with the ephemeris information available from the header file of the imagery. The accuracies of geometrically corrected images using a single ground control point are similar with those of images using many ground control points. The RMSE(root mean square error) of a single ground control imagery using the developed algorithm is about 3.1 pixels and the RMSE of many ground points are 3.89 pixels. Accordingly our result is very applicable to the condition when rapid processing is needed or when it is difficult to acquire many ground control points.

SAR(Synthetic Aperture Radar) 영상의 경우 광학영상과는 달리 지모와 지물에 대한 판별력이 떨어져 기준점의 선택이 매우 어렵다. 또한 SAR영상은 홍수, 지진, 산불 등과 같은 자연재해와 관련된 분야에 많이 활용되고 있어 실시간적 처리가 종종 요구되며, 이때 기준점의 취득이 처리시간에 많은 영향을 미친다. 따라서 본 연구에서는 기준점 선정에 소모되는 시간과 노력을 줄이기 위해서 RADARSAT SAR영상의 천체력자료만을 이용하여 궤도를 모형화하고, 이때 발생하는 영상의 기하학적 오차를 단일 기준점만을 이용하여 엄밀기하보정을 수행하는 기법을 연구하였다. 본 연구의 단일기준점을 이용한 보정기법을 통해 생성된 기하보정영상의 정확도를 평가한 결과 RMSE가 평균 3.1 영상소를 보였으며, 이는 다수의 기준점과 상용소프트웨어인 ERDAS IMAGINE을 이용하여 생성한 기하보정영상(RMSE: 3.89 영상소)과 유사한 정확도를 보였다. 따라서 본 연구의 결과는 기준점의 취득이 난해하거나 신속한 기하보정의 처리가 요구되는 상황에서 효율적인 활용이 가능할 것으로 판단된다.

Keywords

References

  1. IEEE Trans. on Geosci. and Rem. Sens. v.GE-20 Location of Spacebome SAR imagery Curlander, J.C.
  2. IEEE Trans. on Geosci. and Rem. Sens. v.GE-22 Utilization of SAR Data for Mapping Curlander, J.C.
  3. Int. J. Rem. Sens. v.8 A Post-Processing System for Automated Rectification of Spacebome SAR Imagery Culander, J.C.;Kwok, R.;Pang, S.S.
  4. IEEE Trans. on Geosci. and Rem. Sens. v.GE-24 Multiple Incidence Angle SIR-B Experiment Over Argentina: Generation of Secondary Image Products Domik, G.;Leberl, F.W.;Cimino, J.B.
  5. PE & RS v.5 Rectification of terrain induced distortion in radar imagery Kwok, R.;Curlander, J.;Pang, S.S.
  6. Photogrammetric Reduction and Analysis of Real and Simulated SPOT Imageries Makki, S.A.
  7. Astronomy on the Personal Computer Montenbruk, O.;Pfleger, T.
  8. The Rectification of SAR image data using a digital elevation model and image simulation techniques. Phase-B study for ERS-l processing and arching facility, Forschungsgesellschaft Joanneum, Graz, Austriz, ESA Contract Rep. 6292/85/HGE-l, Tech. Note 17. Raggam, J.;Strobl, D.;Triebnig, G.
  9. RADARSAT Data Products Specifications RSI
  10. Satellite Geodesy: Foundation, Methods and Applications Seeber, G.
  11. High accuracy data fusion of satellite and ariphoto images, ACSM/ASPRS Annual Convention & Exposition. Toutin, T.;Cheng, P.
  12. IEEE Trans. on Geosci. and Rem. Sens. v.37 no.5 Error Tracking of Radargrammetric DEM from RADARSAT Images Toutin, T.
  13. IGARSS 2002 Path Processing and Block Bundle Adjustment With RADARSAT-l SAR Images Toutin, T.;Carbonneau, Y.;Chenier, R.