Conference Paper

Arc of ellipse detection for video image registration

Lab. des Sci. de l'Inf. et des Syst., Marseille, France
DOI: 10.1109/SIPS.2005.1579894 Conference: Signal Processing Systems Design and Implementation, 2005. IEEE Workshop on
Source: IEEE Xplore


In this paper, we give a detailed presentation of a robust algorithm for detecting arcs of ellipse in a binary image. The characterization of such arcs of ellipse enables the identification between some video image elements and the corresponding landmarks in a 3D model of the scene to be represented. This algorithm is based on a classical ellipse property that enables its parameters separation. It provides interesting results even in noisy images or when these arcs are small and partially hidden.

Download full-text


Available from: Arnaud le troter, Jan 19, 2016
  • Source
    • "For the same reason, arc detection is often used in addition to line detection. This enables reliable field registration for center regions, even if just the center circle and line is visible [7] [8]. In fact, Wang et al. rely entirely on the detection of arcs, as their system detects the arc of the penalty box as well [9]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, a system for temporally consistent registration of a soccer field is proposed. It is implemented using 2D field line and arc detection as well as motion estimation for frames where too few field lines are detected. We show that a combination of these technologies yields more consistent results than common state-of-the-art approaches. As an important additional contribution, a Kalman filter is applied to suppress remaining misdetections. The system is able to register the soccer field with a mean error of 2.53 m regardless of the visible field region.
    Full-text · Conference Paper · Sep 2013
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Imitating human ringer tip, an arrayed type tactile sensor comprising four layers is under development. In this report, a recognizing method of force and its direction is proposed by using two stage neural networks, and its validity is investigated by simulation. The procedure of simulation is as follows: first, the stress distribution or this sensor sheet applied force is simulated by FEM (Finite Element Method). The obtained stress data are assigned to each assumed stress sensing element of the array. Second, all data or these elements are processed by the first stage neural network, and force direction is calculated. Third, using this direction data and maximal stress data of 4 layers are processed by the second stage neural network.
    Preview · Conference Paper · Jan 2005
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present a global approach that enables the production of 3D soccer sequences from non-calibrated video cameras. Our system can produce a 3D animated model of the scene from a single non-calibrated moving camera (a TV sequence for example). The results presented here are very encouraging even with a single camera approach and will probably improve with the future introduction of multiple images that will help resolving occlusion issues and integrating into a single model information coming from various locations on the field. The key point of our approach is that it doesn’t need any camera calibration and it still works when the camera parameters vary along the process. Details on the registration and tracking processes are given as well as the description of the “Virtual Reality” system used for displaying the resulting animated model.
    Full-text · Conference Paper · Jan 2007
Show more