DOI QR코드

DOI QR Code

특징점 기반 단안 영상 SLAM의 최적화 기법 및 필터링 기법 성능 분석

Performance Analysis of Optimization Method and Filtering Method for Feature-based Monocular Visual SLAM

  • Jeon, Jin-Seok (School of Electrical and Electronics Engineering, ChungAng University) ;
  • Kim, Hyo-Joong (School of Electrical and Electronics Engineering, ChungAng University) ;
  • Shim, Duk-Sun (School of Electrical and Electronics Engineering, ChungAng University)
  • 투고 : 2018.12.11
  • 심사 : 2018.12.23
  • 발행 : 2019.01.01

초록

Autonomous mobile robots need SLAM (simultaneous localization and mapping) to look for the location and simultaneously to make the map around the location. In order to achieve visual SLAM, it is necessary to form an algorithm that detects and extracts feature points from camera images, and gets the camera pose and 3D points of the features. In this paper, we propose MPROSAC algorithm which combines MSAC and PROSAC, and compare the performance of optimization method and the filtering method for feature-based monocular visual SLAM. Sparse Bundle Adjustment (SBA) is used for the optimization method and the extended Kalman filter is used for the filtering method.

키워드

DHJGII_2019_v68n1_182_f0001.png 이미지

그림 1 월드좌표계와 카메라좌표계의 관계 Fig. 1 World coordinate and camera coordinate

DHJGII_2019_v68n1_182_f0002.png 이미지

그림 2 RANSAC(좌)과 MPROSAC(우) 알고리즘 Fig. 2 RANSAC(left) and MPROSAC(right) algorithm

DHJGII_2019_v68n1_182_f0003.png 이미지

그림 3 MPROSAC 사용 전(좌)과 MPROSAC 사용 후(우)Fig. 3 Before (left) and after (right) using MPROSAC

DHJGII_2019_v68n1_182_f0004.png 이미지

그림 4 2D-2D 카메라 자세 추정 알고리즘 블록도 Fig. 4 Block diagram for 2D-2D camera pose estimation

DHJGII_2019_v68n1_182_f0005.png 이미지

그림 5 3D-2D 카메라 자세 추정 알고리즘 블록도 Fig. 5 Block diagram for 3D-2D camera pose estimation

DHJGII_2019_v68n1_182_f0006.png 이미지

그림 6 BA 알고리즘 블록도 Fig. 6 Block diagram for BA algorithm

DHJGII_2019_v68n1_182_f0007.png 이미지

그림 7 확장 칼만필터 알고리즘 블록도 Fig. 7 Block diagram for Extended Kalman filter algorithm

DHJGII_2019_v68n1_182_f0008.png 이미지

그림 8 2D-2D 자세 추정방법 결과 Fig. 8 Result of 2D-2D camera pose estimation

DHJGII_2019_v68n1_182_f0009.png 이미지

그림 9 Scale 보정 후 2D-2D 자세 추정방법 결과 Fig. 9 Result of 2D-2D camera pose estimation after scale correction

DHJGII_2019_v68n1_182_f0010.png 이미지

그림 10 3D-2D 자세 추정방법 결과 Fig. 10 Result of 3D-2D camera pose estimation

DHJGII_2019_v68n1_182_f0011.png 이미지

그림 11 3D-2D 기반 번들조정 후 3차원 점 성능 결과 Fig. 11 Result of 3D points after 3D-2D based BA

DHJGII_2019_v68n1_182_f0012.png 이미지

그림 12 확장 칼만필터 3차원 점 성능 결과 Fig. 12 Result of 3D points after EKF

표 1 RANSAC, MSAC, PROSAC, MPROSAC 성능비교 Table 1 Performance comparison among RANSAC, MSAC, PROSAC, and MPROSAC

DHJGII_2019_v68n1_182_t0001.png 이미지

참고문헌

  1. H. Strasdat, J. M. M. Montiel, A. J. Davison, "Visual SLAM: Why filter?", Image and Vision Computing 30, pp. 65-77, 2012. https://doi.org/10.1016/j.imavis.2012.02.009
  2. Sunglok Choi, Taemin Kim, Wonpil Yu, "Performance evaluation of RANSAC family", British Machine Vision Conference, BMVC 2009, London, UK, pp. 81.1-81.12, 7-10 September 2009.
  3. Herbert Bay, Tinne Tuytelaars, Luc Van Gool, "SURF: Speeded Up Robust Features", Computer Vision and Image Understanding, Vol. 110, Issue 3, pp. 346-359, June 2008. https://doi.org/10.1016/j.cviu.2007.09.014
  4. Martin A. Fischler and Robert C. Bolles, "Random Sample Consensus: A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography", Communications of the Association for Computing Machinery, Vol. 24, No. 6 pp. 381-395, June 1981. https://doi.org/10.1145/358669.358692
  5. P. H. S. Torr, A. Zisserman, "MLESAC: A new robust estimator with application to estimating image geometry", Computer Vision and Image Understanding, Vol. 78, pp. 138-156, 2000. https://doi.org/10.1006/cviu.1999.0832
  6. Richard I. Hartley, "In Defense of the Eight-Point Algorithm", IEEE Transactions on Pattern Recognition and Machine Intelligence, Volume 19, (6), pp. 580-593, June 1997. https://doi.org/10.1109/34.601246
  7. MANOLIS I. A. LOURAKIS and ANTONIS A. ARGYROS, "SBA: A Software Package for Generic Sparse Bundle Adjustment", ACM Transaction on Mathematical Software(TOMS), Vol. 36, No 1, Article 2, pp. 2:1-2:30, March 2009.
  8. M.E. Ragab, and K.H. Wong, "Extended Kalman Filter Based Pose Estimation Using Multiple Cameras", The CSE Dept., The Chinese University of Hong Kong, Internal report, 16 May, 2007.
  9. Jin-Seok Jeon, Hyo-Joong Kim, Duk-Sun Shim, MSAC/PROSAC Fusion Algorithm to Enhance SURF Performance, Conference on Information and Control Systems, pp. 276-277, 26 Oct, 2018.