Z. Lei, T. Tasdizen, and D.B. Cooper. Pims and invariant parts for shape recognition. In Proceedings of Sixth International Conference on Computer Vision, pages 827-- 832, Mumbai, India, 1998. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Improving The Stability Of Algebraic Curves For Applications - Tasdizen, Tarel, Cooper (2000) (2 citations) Self-citation (Tasdizen Cooper) (Correct)
....Acknowledgment This work was partially supported by NSF Grant #IIS 9802392. 1 1 Introduction Algebraic 2D curves (and 3D surfaces) are extremely powerful for shape recognition and singlecomputation pose estimation because of their fast tting, invariants, and interpretable coecients, [10, 11, 13, 17, 18, 19, 21]. Signi cant advantages over Fourier Desciptors are their applicability to non star shapes, to open curves, to curves that contain gaps, and to unordered curve data, Sec. 2. Under circumstances where these issues are not relevant, polynomials based on Fourier analysis may be very e ective, and an ....
....9 for examples where k was chosen in this manner. This can be done iteratively since tting for modeling can usually be done o line. Parameter k can be increased from 0 to larger values until signi cant amounts of error start to be introduced into the t. Polynomial Interpolated Measure (PIM) [11] can be used to track this error as a di erence in the polynomial at k = 0 and at the value of k under consideration. Choosing k for Recognition. Here the main goal is to minimize the total mean squared error of estimator A rr . Such an optimal value of k is empirically shown to exist and is ....
Z. Lei, T. Tasdizen, and D. Cooper. Pims and invariant parts for shape recognition. In Proceedings of Sixth International Conference on Computer Vision (ICCV'98), pages 827-832, Mumbai, India, 1998. 25
The Complex Representation of Algebraic Curves and its Simple.. - Tarel, Cooper (2000) Self-citation (Cooper) (Correct)
....invariant recognition, and then do more careful comparison on the shapes that remain. This more careful comparison would involve pose estimation for alignment followed by careful comparison of aligned shape data. This careful comparison of aligned shapes could then be done through our PIMs measure [10] or through other measures. How do Fourier descriptors compare with the algebraic curve model Fourier descriptors, like algebraic curves, provide a global description for shapes from which pose and recognition can be processed. But the Fourier approach has difficulty in general with open patches ....
.... and 3D surfaces, the most basic approach to comparison of two shapes is iterative estimation of the transformation of one algebraic model to the other followed by recognition based on comparison of their coefficients or based on comparing the data set for one with the algebraic model for the other [12, 1, 10]. But the problem of initialize this iterative process still remains. A major jump was the introduction of intrinsic coordinate systems for pose estimation and Euclidean algebraic invariants for algebraic 2D curves and 3D surfaces [1, 6] These are effective and useful, but as published do not use ....
[Article contains additional citation context not shown here]
Z. Lei, T. Tasdizen, and D.B. Cooper. Pims and invariant parts for shape recognition. In Proceedings of Sixth ICCV, pages 827--832, Mumbai, India, 1998. also as LEMS Tech. Report 163, Brown University.
PIMs and Invariant Parts for Shape Recognition - Lei, Tasdizen, Cooper (1997) (1 citation) Self-citation (Lei Tasdizen Cooper) (Correct)
....(Polynomial Interpolated Measure) which compares the distance between data sets in terms of the distance between their fitted polynomial coefficients weighted in a certain way. It provides a new concept for comparing data curves and leads to many useful results (e.g. orthogonal decomposition [8] ) Consider data sets Z 1 and Z 2 containing N 1 and N 2 points, respectively. The fitted IP (implicit polynomial) curve models for these are f 1 (x; y) P 0i;j;i j4 a 1ij x i y j and f 2 (x; y) P 0i;j;i j4 a 2ij x i y j (for simplicity we use 4th degree IPs here) Define the ....
Z. Lei, T. Tasdizen, and D. Cooper, "PIMs and Invariant Parts for Shape Recognition," LEMS Tech. Report 163, Division of Engineering, Brown University, 1997.
Kernel Based Algebraic Curve Fitting - Christian Walder Brian (Correct)
No context found.
Z. Lei, T. Tasdizen, and D.B. Cooper. Pims and invariant parts for shape recognition. In Proceedings of Sixth International Conference on Computer Vision, pages 827-- 832, Mumbai, India, 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