S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. PAMI, 15(2):136--144, February 1993.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the correspondence problem. In remote sensing applications, point pattern matching is used for image registration. For the general problem of point pattern matching, where no a priori knowledge about the two sets of points is available, a number of algorithms have been described in the literature [17, 222, 205, 9, 221, 217]. Baird s O(n 2 ) algorithm, where n is the number of points in each of the two point sets, becomes more complex (O(n 3 ) when the number of points in the two sets are not the same. Vinod et al. 222] propose a neural network for point pattern matching after formulating it as a 0 1 integer ....

S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. IEEE Trans. on Pattern Analysis and Machine Intelligence, 15(2):136--144, February 1993.

....by examining the e ectiveness of the recognition system on a variety of examples. 1 Introduction The representation of the shape of objects can have a signi cant impact on the effectiveness of a recognition strategy. Shapes have been represented as curves [12, 22, 15, 2, 7, 23] point sets [1, 16, 21], feature sets [3, 8] and by medial axis [24, 18, 19, 14, 13, 11, 10] among others. This paper develops an approach to object recognition based on a curve based representation of shape outline using the proposed concept of an alignment curve, and identi es the strengths and weaknesses of using ....

....recognition and for indexing into image databases by shape context. In many object recognition and content based image indexing applications, the object outlines are represented as curves and matched. The matching relies on either aligning feature points using an optimal similarity transformation [1, 16, 21] or on a deformation based approach to aligning the properties of the two curves [12, 22, 15, 2, 7, 6, 23] Deformation based methods typically involve nding a mapping from one curve to the other that minimizes an elastic performance functional, which penalizes stretching and bending [4, 20, ....

[Article contains additional citation context not shown here]

S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. PAMI, 15(2):136-144, February 1993.

....standard approach taken in computational geometry is the development of exact, provably correct and efficient solutions to geometric problems. First some related work is mentioned in the next subsection. 1. 1 Related work Matching has been approached in a number of ways, including tree pruning [55], the generalized Hough transform [8] or pose clustering [51] geometric hashing [59] the alignment method [27] statistics [40] deformable templates [50] relaxation labeling [44] Fourier descriptors [35] wavelet transform [31] curvature scale space [36] and neural networks [21] The ....

S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. IEEE Transactions on Pattern Analysis and Machine Intelligence, 15(1):136--144, 1993.

....traditional search methods. When a large number of models is needed and an incremental development of the models is required, then this method may be very beneficial. 2. 11 Detection of Classes of Objects Another technique that can be used to detect classes of objects is parameter point matching [13]. With this method, objects whose parts may move with respect to one another may be classified as well as objects with various shapes and proportions that can Page 1212 be classified into one parametric model. A parameterized point is made up of a location vector, direction, and deformation ....

Umeyama, S., (1993). Parameterized Point Pattern Matching and Its Application to Recognition of Object Families. IEEE Trans. on Pattern Analysis and Machine Intell., Vol. 15, No. 2, pp. 136 - 144.

....to solve the correspondence problem. In remote sensing applications, point pattern matching is used in image registration. For the general case of point pattern matching, where no a priori knowledge about the two sets of points is available, many algorithms have been described in the literature [1, 2, 3, 4, 5, 6]. Baird s O(n 2 ) algorithm, where n is the numbers of points in each of the two point sets, becomes more complex when the number of points are not same in the two sets. Vinod et al. 2] propose a neural network for point pattern matching after formulating the problem as a 0 1 integer ....

S. Umeyama, "Parameterized point pattern matching and its application to recognition of object families," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 15, pp. 136--144, February 1993.

....representation of patterns, the transformation group, and the dissimilarity measure. Figure 2: Fingerprint matching. Figure 3: Query hieroglyph (left) and hieroglyphs retrieved from database, from [VV99] 2 Approaches Matching has been approached in a number of ways, including tree pruning [Ume93] the generalized Hough transform or pose clustering [Bal81] Sto87] geometric hashing [WR97] the alignment method [HU87] statistics [Sma96] deformable templates [SP95] relaxation labeling [RR80] Fourier descriptors [Lon98] wavelet transform [JFS95] curvature scale space [MAK96] and ....

S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. IEEE Transactions on Pattern Analysis and Machine Intelligence, 15(1):136-144, 1993.

....matching. While the correspondence problem has a long history in rigid, affine and projective point matching [14] there is relatively a dearth of literature on non rigid point matching. Recently, there has been some interest in using point based correspondence strategies in non rigid matching [23, 10, 29, 32, 21]. The modal matching approach in [23] relies on the point correspondence approach pioneered in [24] and further developed in [25] The basic idea here is to use a pairing matrix that is built up from the Gaussian of the distances between any point feature in one set and the other. The modes of ....

S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. IEEE Trans. Patt. Anal. Mach. Intell., 15(1):136--144, Jan. 1993.

....While enormous amount of literature exists in each of the fields, these works have typically been developed in isolation. The following is a representative list: contour modeling [13, 33, 35, 48] edge detection [8, 11, 17, 30, 43] edge linking [3, 32, 34] contour detection and classification [18, 20, 36, 45]. This strict dichotomy of visual tasks has been widely adopted because it decomposes the problem of image analysis into independent and managable components. Unforturnately, the Marr s paradigm also results in unidirectional and irreversible flow of error from one component to the next. Moreover, ....

.... active contour model (snake) 21] parameterized templates [47] elliptic Fourier decompositions [41] implicit polynomials [42] handprinted character recognition [19] and affine invariant contour tracking [5] In addition, several methods that classify deformable contours have also been reported [18, 45], but these ignore the problem of extraction. There have been relatively few methods that consider detection and classification of deformable contours directly from noisy images [19, 42] 1.2 Contributions This thesis presents an integrated approach in modeling, extracting, detecting and ....

[Article contains additional citation context not shown here]

S. Umeyama, "Parameterized Point Pattern Matching and Its Application to Recognition of Object Families," IEEE Trans. Pat. Anal. Mach. Intell., vol. PAMI-15, 1993, pp. 136-144.

....projection. Each vertex edge pair casts a vote in transformation space for a particular transformation with the true one(s) estimated as peaks in this space. Finally, we briefly mention the relatively recent developments of recognition through the use of projective invariants, pioneered by Weiss [39], and best illustrated in the system developed by Rothwell et al. [14,35] To date most of this work has concentrated on planar objects. Only recently [34] have extensions to three dimensional objects been achieved (for objects which can be caged by polyhedral point sets, and objects with ....

S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. IEEE Transactions on Pattern Analysis and Machine Intelligence, 15(2):136--144, February 1993.

....and similarities between objects are measured in terms of modal deformation energies. A limitation of this technique is in its ability to handle partial matches or to match in the presence of occlusion. Also, work has been done on parameterized parts using an exponential time tree search technique [Umeyama, 1993], and on local affine transforms [Wakahara, 1994] but neither explicitly considers the non rigid matching, expressing a one to one correspondence between points. Hel Or and Werman, 1994] lay out a framework for characterizing an articulated object, indicating constraints in the relative ....

Umeyama, S. (1993). Parameterized point pattern matching and its application to recognition of object families. IEEE Transactions on Pattern Analysis and Machine Intelligence, 15:136--144.

....of the correspondence is improved by the integration of multiple features. A method for non rigid parameter estimation is developed, based on the assumption of a well articulated model with distinct regions, each of which may move in an affine fashion, or can be approximated as such. Umeyama [3] has done work on parameterized parts using an exponential time tree search technique, and Wakahara [4] on local affine transforms, but neither integrates multiple features nor explicitly considers the non rigid matching case, while expressing a one to one correspondence between points. 2 Affine ....

S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. IEEE Trans. on Pattern Analysis and Machine Intelligence, 15:136--144, 1993.

....demonstrate the robustness of the algorithm. 2 Related Work A large number of different approaches have been tried on point matching. Tree pruning methods involve searching over a tree of possible matches while eliminating portions of the search space (Baird, 1984; Grimson and Lozano Perez, 1987; Umeyama, 1993). The generalized Hough transform requires the division of the parameter space of possible poses into discrete bins wherein good matches are registered as votes in the appropriate bin (Ballard, 1981; Stockman, 1987) Geometric hashing is another voting scheme where discrete bins are created for ....

Umeyama, S. (1993). Parameterized point pattern matching and its application to recognition of object families. IEEE Trans. Patt. Anal. Mach. Intell., 15(1):136--144.

....These can be classified under i) tree searches, ii) Hough transforms, iii) geometric hashing and iv) the alignment method. Tree search methods involve searching over a tree of possible matches while eliminating branches of the tree from consideration (Baird, 1984; Grimson and Lozano Perez, 1987; Umeyama, 1993). The generalized Hough transform divides the valid set of spatial mappings into discrete bins. Good matches are registered as votes in the appropriate bin (Ballard, 1981; Stockman, 1987) In geometric hashing, depending on the spatial mapping, a set of possible bases that can represent the point ....

Umeyama, S. (1993). Parameterized point pattern matching and its application to recognition of object families. IEEE Trans. Patt. Anal. Mach. Intell., 15(1):136--144.

....primitive matching, and syntactic matching. The idea behind proximity matching methods is to search for the best matching while permitting the rotation, translation and scaling (to be called alignment transformation) of each curve, such that the distances between matched key points are minimized [4, 22, 24, 43]. Consequently these methods are rather slow; moreover, if scaling is permitted an erroneous shrinking of one feature set may result, followed by the matching of the entire set with a small number of features from the other set. One may avoid these problems by excluding many toone matches and by ....

Umeyama S., "Parameterized point pattern matching and its application to recognition of object families ", PAMI 15, 136-144, 1993.

....boundary extraction into Bayesian posterior estimation and thus exhibits high level of robustness and adaptability. This paper demonstrates the usefulness of g snake in classifying among several candidate deformable contours directly from noisy images. Although there already exist numerous methods [7, 8] that classify deformable contours, these typically assume that the contour has been previously extracted. However, without a contour model, contour extraction has been shown to be an illposed problem [9] Thus the practicality of these methods is limited in applications involving noisy images. ....

S. Umeyama, "Parameterized Point Pattern Matching and Its Application to Recognition of Object Families," IEEE Trans. Pat. Anal. Mach. Intell., vol. PAMI-15, 1993, pp. 136-144.

No context found.

S. Umeyama. Parameterized point pattern matching and its application to recognition of object families. PAMI, 15(2):136--144, February 1993.

No context found.

S. Umeyama, "Parameterized point pattern matching and its application to recognition of object families," IEEE Transaction on pattern Analysis and Machine Intelligence, 15(1):136-144,1993.

No context found.

Umeyama S., \Parameterized point pattern matching and its application to recognition of object families ", PAMI 15, 136-144, 1993.

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