| JI, Q., COSTA,M.,HARALICK,R.,AND SHAPIRO, L. An integrated linear technique for pose estimation from different features. In International Journal of Pattern Recognition and Artificial Intelligence(1999), June. |
.... Other work that explicitly uses ellipses as part of planar patterns for calibration and pose estimation include Rothwell [6] who calculates a set of specific points using tangent lines and intersection, Song [7] who uses an iterative approach that requires a corresponding pair of conics, and Ji [3] who considers ellipses as one of many geometric primitives (the others are points and lines) and gives a least squares solution for arbitrary sets of matched primitives. A planar pattern s 2D features can be di#cult to extract from oblique camera angles (below 30 degrees) This means that one ....
Q. Ji, M. Costa, R. Haralick, and L. Shapiro. An integrated linear technique for pose estimation from di#erent features, 1999.
....Song [18] who uses an iterative approach and requires a corresponding pair of conics [16] and Kaminski [12] who reconstructs the conic and the calibration parameters. Most related to our solution is the single image direct methods of computing planar pose. Much of this work is summarized in [11], which integrates previous work into a unified geometric framework which uses combinations of point and line features. Spatially aware devices: Non traditional user interfaces, such as Graspable [6] Tangible [10] or Manipulable [8] interfaces, have been shown to be better than the traditional ....
Q. JI, M. COSTA, R. HARALICK, and L. SHAPIRO. An integrated linear technique for pose estimation from di#erent features, 1999.
....from points, lines and curves has been extensively studied. Estimation based only on point correspondences from four [25] and more points is introduced in [13, 19] Haralick, et.al. 20] provides a review of many 3 point techniques with a careful analysis of their stability. Qiang, et.al. [39] develops an analytic least squares technique for pose estimation from points, lines and ellipse circle pairs. These methods all assume a known correspondence between geometry and image features, which often requires extensive user involvement. Several techniques have been developed for automatic ....
....## # # ## ## # # ## #### ## # # ## # # # # # using a standard least squares optimizer. As this optimization is typically prone to a small number of local minima, we run it from multiple initial positions. A detailed treatment of pose estimation from circle ellipse pairs can be found in [39]. Since this system with 10 equations and 6 unknowns is overdetermined, each solution returns an error that can be used to weed out most bad correspondences immediately. Otherwise, we use this initial guess for the pose to find a small set (typically less than 3) of candidate correspondences for ....
JI, Q., COSTA, M., HARALICK, R., AND SHAPIRO, L. An integrated linear technique for pose estimation from different features. In International Journal of Pattern Recognition and Artificial Intelligence(1999), June.
No context found.
JI, Q., COSTA,M.,HARALICK,R.,AND SHAPIRO, L. An integrated linear technique for pose estimation from different features. In International Journal of Pattern Recognition and Artificial Intelligence(1999), June.
No context found.
Q. Ji, M. S. Costa, R. M. Haralick and L. G. Shapiro, An Integrated Linear Technique for Pose Estimation from Different Geomet ric Features, Technical Report, University of Washington, Department of Electrical and Computer Engineering, 1998.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC