TY - JOUR
T1 - Optimal matching of closed contours with line segments and arcs
AU - Chen, Jen Ming
N1 - Funding Information:
Recently, the shape matching problems have been formulated as optimization models and solved by sophisticated mathematical techniques, such as using the Lagrangian relaxation and the primal-dual schemes (Mundy et al., 1992; Noble et al., 1992; Ventura et al., 1995) and the gradient-based descent algorithms (Chen J This research was partially supported by the National Science Council (Taiwan) under Grant NSC84-2213-E-008-011. 2 E-mail: [email protected].
PY - 1997/6
Y1 - 1997/6
N2 - In this paper, a planar contour consisting of line segments and circular arcs is described in analytic form, and the contour matching problem is formulated as an optimization problem based on the criterion of minimax (or so-called minimal zone) errors. An iterative descent algorithm is developed to solve the multimodal and non-differentiable optimization problem, the solution of which reports the out-of-profile error and determines the scale and pose of the shape.
AB - In this paper, a planar contour consisting of line segments and circular arcs is described in analytic form, and the contour matching problem is formulated as an optimization problem based on the criterion of minimax (or so-called minimal zone) errors. An iterative descent algorithm is developed to solve the multimodal and non-differentiable optimization problem, the solution of which reports the out-of-profile error and determines the scale and pose of the shape.
KW - Decent algorithm
KW - Minimax error
KW - Optimization
KW - Shape representation
UR - http://www.scopus.com/inward/record.url?scp=0031168576&partnerID=8YFLogxK
U2 - 10.1016/S0167-8655(97)00040-8
DO - 10.1016/S0167-8655(97)00040-8
M3 - 期刊論文
AN - SCOPUS:0031168576
SN - 0167-8655
VL - 18
SP - 567
EP - 574
JO - Pattern Recognition Letters
JF - Pattern Recognition Letters
IS - 6
ER -