Information extraction of the moving objects based on edge detection and optical flow

Edge 검출과 Optical flow 기반 이동물체의 정보 추출

  • 장민혁 (조선대학교 전자${\cdot}$정보통신공학부 DSP&멀티미디어 연구실) ;
  • 박종안 (조선대학교 전자${\cdot}$정보통신공학부 DSP&멀티미디어 연구실)
  • Published : 2002.08.01

Abstract

Optical flow estimation based on multi constraint approaches is frequently used for recognition of moving objects. However, the use have been confined because of OF estimation time as well as error problem. This paper shows a new method form effectively extracting movement information using the multi-constraint base approaches with sobel edge detection. The moving objects anr extraced in the input image sequence using edge detection and segmentation. Edge detection and difference of the two input image sequence gives us the moving objects in the images. The process of thresholding removes the moving objects detected due to noise. After thresholding the real moving objects, we applied the Combinatorial Hough Transform (CHT) and voting accumulation to find the optimal constraint lines for optical flow estimation. The moving objects found in the two consecutive images by using edge detection and segmentation greatly reduces the time for comutation of CHT. The voting based CHT avoids the errors associated with least squares methods. Calculation of a large number of points along the constraint line is also avoided by using the transformed slope-intercept parameter domain. The simulation results show that the proposed method is very effective for extracting optical flow vectors and hence recognizing moving objects in the images.

다제약 접근기반 OF(optical flow) 평가기술이 이동 물체의 인식에 자주 이용되고 있다. 그러나 OF 평가시간 뿐만 아니라 오차 문제로 인하여 사용이 제한되고 있다. 본 논문에서는 sobel 에쥐 검출과 다제약 접근기반 OF를 이용하여 효율적으로 움직임 정보를 추출하는 방법을 제안한다. 먼저 에쥐 검출 후 차영상과 영역분할기법으로 영상열 내 이동물체를 검출하고 임계치 처리로 잡음에 의해 검출된 이동물체들을 제거한다. 그리고 OF 최적 제약선을 찾기 위한 CHT와 Voting 누적을 적용한다. 이때 에쥐 검출과 영역분할을 이용함으로써 연속하는 영상열 내에서 이동 물체를 찾기 위한 CHT 계산시간을 현저히 줄이는 것이 가능하다. CHT 기반의 Voting은 최소자승법을 가미함으로써 오차 또한 감소시킨다. 그리고 제약선에 따른 수많은 점들을 계산하는 작업도 변환된 기울기-교점 파라미터를 사용함으로써 줄어들게 된다. 시뮬레이션 결과 영상 내에서 이동물체 인식비가 증가됨을 보였고 이동물체의 움직임 정보를 제공하는 OF 벡터도 매우 효율적으로 검출됨을 확인하였다.

Keywords

References

  1. K. Prazdny, 'On the information in opticalflow', Computer Vtsion Graphics Image Process, 23, pp.239-59, 1983 https://doi.org/10.1016/0734-189X(83)90025-7
  2. P. J. Burt, et al., 'Object tracking with a movingcamera', Proc. IEEE Workshop on VisuatMotion, hvine, Califoniia, U.S.A., pp.2-12,1989
  3. R. C. Nelson and J. Aloimonos, 'Obstacleavoidance using field divergence', IEEE Trans.Pattern AnaIysis Mach. Intett. 11(10), pp. 1102-1106, 1989 https://doi.org/10.1109/34.42840
  4. M. Subbarao, 'Bounds on time-to-collision androtation component from first-order derivativesof image flow', Computer Vision GraphicsImage Process, 50, pp.329-341, 1990 https://doi.org/10.1016/0734-189X(90)90151-K
  5. B. K. P. Hom and B. G. Schunck, 'Determiningoptical flow', Artif Intrll. 17, pp.185-203, 1981 https://doi.org/10.1016/0004-3702(81)90024-2
  6. H. H. Nagel, 'Displacement vectors derivedfrom second-order intensity variations in imagesequences', Computer Vision Graphics Image Process, 21, pp.85-117, 1983 https://doi.org/10.1016/S0734-189X(83)80030-9
  7. H. H. Nagel and W. Enkelmann, 'Towards theestimation of displacement vector fields by'oriented smoothness' constraints', Proc. 7thIEEE Int. Co on Pattern Recognition, pp.6-8,1984
  8. C. Cafforio and P. Rocca, 'Tracking movingobjects in television images', Sienat Process, 1,pp., 1979 https://doi.org/10.1016/0165-1684(79)90015-X
  9. M. Campani and A. Verri, 'Computing opticalflow from an overconstrained system of linearalgebraic equations', Proc. 3rd IEEE Int. Conf.on Computer Vision, pp.22-26, 1990
  10. P. Nesi, A. Delbimbo and J. L. Sanz,'Multiconstraints-based optical flow estimationand segmentation', Int. Workshop on ComputerArchitecture for Machine Perception, Paris,pp.419-426, 1991
  11. W. B. Thompson, 'Combining motion andcontrast for segmentation', IEEE Trans. PAMI,2(6), pp.543-549, 1980
  12. S. Gemen et. al., 'stochastic relaxation, Gibbsdisthbution, and Bayesian restoration ofimages', IEEE Trans PAMl, 6(6), pp.721-741,1985
  13. M. J. Black, 'Combining intensity and motionfor incremental segmentation and tracldng overlong image sequence', Proc. ECCV92,pp.485-493, 1992
  14. M. Etoh et. al., 'Segmentation and 2D motionestimate by region fragments', Proc. 4th Int.Co Computer Vision, pp.192-199, 1993
  15. D. Ben-Tzvi and M. Sandler, 'A combmatorialHough transfonn', Pattem Recognition, Lett. 11,pp.167-174, 1990 https://doi.org/10.1016/0167-8655(90)90002-J
  16. Sung-Kwan Kang, Jong-An Park, Estimation ofMoving Infonnation for Tracking of MovingObjects, The Korean Society of MechanicalEngineers, Vol. 15, No. 3, March 2001