L. J. Latecki and R. Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. CVIU, 73(3):441--454, 1999.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....of shape hierarchy leads to a multi scale shape decomposition approach 25 based on discrete curve evolution to yield robust computation and hierarchical organization of parts. Multi scale shape decomposition is based on curve evolution obtained from different operators. Latecki and Lakmper [54] proposed curve evolution by linearization. A continuous curve is first decomposed into maximal digital line segments. Then the evolution proceeds by substituting two consecutive line segments with a single line segment joining their endpoints. However this approach is very sensitive to the order ....

....body part shapes and geometric relationships. Zhu and Yuille [23] developed a similarity measure to compare silhouettes based on both the local shapes of parts and the topology but the method can not handle shape degeneration or resolution changes very well. Several curve evolution approaches [50, 54, 55] have been proposed to model shapes of an object at different scales, but the related similarity measure is sensitive to occlusion and is not invariant under scaling. Leung et al. 27, 28, 29] have proposed a method which combines the intensity pattern and the spatial relationships between the ....

L. J. Latecki and R. Lakmper, "Convexity Rule for Shape Decomposition Based on Discrete Contour Evolution," Computer Vision and Image Understanding (CVIU), Vol. 73, No. 3, pp. 441-454, 1999.

....similar to that byAbeet al... 1] A structuring elementwas applied to the segmented branches of the axes, and these were then merged dependenton their convexityvalue. Again various parameters were required to control the axes segmentation and merging stages. Recently Latecki and Lakamper [8]avoided the many of the difficulties of the above approaches, using their so called discrete evolution by digital linearization . Boundary points are iteratively deleted (or equivalently adjacent line pairs are merged) until the resulting shape is convex. Ateach iteration the line pair merge ....

L.J. Latecki and R Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. Computer Vision and Image Understanding, 73(3):441-- 454, 1999.

....zero crossings continuously move along the contour, until two such positions meet and annihilate. Matching of two objects can be done by matching points of annihilation in the ### ## plane [36] Another way of reducing curvature changes is based on the turning angle function (see section 4. 5) [34]. Matching with the curvature scale space is robust against slight affine distortion, as has been experimentally determined [37] Be careful, however, to use this property for fish recognition, see section 3. 2 Matching problems Shape matching is studied in various forms. Given two patterns and ....

L. J. Latecki and R. Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. Computer Vision and Image Understanding, 73(3):441--454, 1999.

....of convexity and an optimization scheme. Since it uses exhaustive search to find the optimal cuts, the algorithm is slow despite the use of a multi scale approach. Another limitation is that the method requires that the number of cuts for the shape partitioning should be given. Latecki, et al. [12] used a contour evolution method to identify convex parts at different stages. The assumption is that significant visual parts will become convex at higher stages of the evolution. They named this rule the hierarchical convexity rule. In related work, Belyaev, et al. 2] developed a numerical ....

....must be added to our system. Perceptually motivated criteria will be used to determine where how to split regions, based on the local shape properties of the region group s bounding contour. As described in Sec. 2, the convexity rule can be used to evaluate the quality of the cut in shape parsing[12, 16]. In our application, we can also make use of our deformable shape model in guiding the selection of splits. In considering a potential split, the change in the fitting cost (Eq. 3) after splitting provides a measure for the quality of the cut. 4.1 Region Splitting Criteria In order to detect ....

[Article contains additional citation context not shown here]

L.J. Latecki and R. Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. CVIU, 73(3):441--454, 1999.

....of centroid of blocks counted in bin (2, 3) In addition, the first vector component is the frame number so that the feature vector keeps temporal information. Consequently, each feature vector has 148 elements. 4. Sort feature vectors by relevance with respect to the general shape of the curve [7, 6]. This method is described later in this section. 5. Construct histogram using number of frames with relevance greater than r as function of r with ranges normalized to [0; 1] 6. Find relevance level corresponding to the point where the curve simplification summarizes the large scale trends of ....

....linear zooms, because this generates straight trajectories in our feature space. In previous work [1] our curve simplification technique was based on binary curve splitting with the Ramer algorithm. However, we found that a method based on polygon simplification by relevance ranking of vertices [7] provides more compelling results. Relevance is the degree of how important a point is to the general shape of the polygon. Even though the work in [7] depicts the case of planar polygons, the same method can be used to obtain relevance with high dimensional points because it considers only ....

[Article contains additional citation context not shown here]

L. J. Latecki and R. Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. In Computer Vision and Image Understanding, volume 73, pages 441--454, 1999.

....along the contour, until two such positions meet and annihilate. Matching of two objects can be done by matching points of annihilation in the s; plane [MAK96] Another way of reducing curvature changes is based on the turning angle function (see Section 5. 1) or tangent space representation [LL99] 2.3 Voting schemes The voting schemes discussed here generally work on so called interest points. For the purpose of visual information systems, such points are for example corner points detected in images. Geometric hashing [LW88, WR97] is a method that determines if there is a transformed ....

Longing Jan Latecki and Rolf Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. Computer Vision and Image Understanding, 73(3):441-454, 1999.

No context found.

L. J. Latecki and R. Lakamper (1999). Convexity Rule for Shape Decomposition Based on Discrete Contour Evolution. Computer Vision and Image Understanding 73:441--454.

No context found.

L. J. Latecki and R. Lakamper (1999). Convexity Rule for Shape Decomposition Based on Discrete Contour Evolution. Computer Vision and Image Understanding 73:441--454.

No context found.

L. J. Latecki and R. Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. Computer Vision and Image Understanding, 73:441--454, 1999.

....of the maximal convex concave arcs contained in object contours is established. The maximal convex concave arcs are not taken from the original contours, but from their simplified versions. Significant maximal convex arcs on simplified contours correspond to significant parts of visual form [9], whose importance and relevance for object recognition is verified by numerous cognitive experiments [5, 6, 16, 18] For P298 a single simplified contour is used as a shape descriptor. This contour is optimally determined by a novel process of contour simplification called discrete curve ....

L. J. Latecki and R. Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. Computer Vision and Image Understanding, 73:441--454, 1999.

....requirements (1) and (2) Observe that our discrete curve evolution is context sensitive, since whether shape components are relevant or irrelevant cannot be decided without context. Our discrete curve evolution is briefly presented in this section (more detailed presentations are given in [12, 13]) Our curve evolution method does not require any control parameters to achieve the task of shape simplification, i.e. there are no parameters involved in the process of the discrete curve evolution. However, we clearly need a stop parameter, which is the number of iterations the evolution is ....

....function normalized with respect to the total length of a polygonal curve C. The main property of this relevance measure is the following The higher the value of K(s, s2) the larger is the contribution to the shape of the curve of arc s to s2. A cognitive motivation of this property is given in [12], where a detailed description of our discrete curve evolution can also be found. Online demonstrations can be viewed on our www site [14] It is a simple and natural observation that maximal convex parts of objects determine visual parts. The fact that visual parts are somehow related to ....

[Article contains additional citation context not shown here]

L. J. Latecki and R. Lak/tmper. Convexity rule for shape decomposition based on discrete contour evolution. Computer Vision and Image Understanding, 73:441-454, 1999.

....these vertices are the most linear ones. Consequently, the remaining vertices of the simplified polyline are frames that are more non predictable than the deleted ones. Our approach to simplification of video polylines is based on a novel process of discrete curve evolution presented in [6] and applied in the context of shape similarity of planar objects in [7] However, here we will use a different relevance measure of vertices. Fig. 1 illustrates the curve simplification produced by the discrete curve evolution for a planar figure. Notice that the most relevant vertices of the ....

L. J. Latecki and R. Lakamper. "Convexity rule for shape decomposition based on discrete contour evolution," CVIU, 73:441--454, 1999.

....these vertices are the most linear ones. Consequently, the remaining vertices of the simplified polyline are frames that are more non predictable than the deleted ones. Our approach to simplification of video polylines is based on a novel process of discrete curve evolution presented in [6] and applied in the context of shape similarity of planar objects in [7] However, here we will use a different relevance measure of vertices. Fig. 1 illustrates the curve simplification produced by the discrete curve evolution for a planar figure. Notice that the most relevant vertices of the ....

L. J. Latecki and R. Lakamper. "Convexity rule for shape decomposition based on discrete contour evolution," CVIU, 73:441--454, 1999.

....frames. Thus, a necessary condition for a simplification of a video polyline is that the sequence of vertices of the simplified polyline be a subsequence of the original one. Our approach to simplification of video polylines is based on a novel process of discrete curve evolution presented in [9] and applied in the context of shape similarity of planar objects in [11] However, here we will use a different measure of the relevance of vertices, described below. Aside from its simplicity, the process of discrete curve evolution differs from the standard methods of polygonal approximation, ....

Latecki, L.J., and Lakamper, R., "Convexity Rule for Shape Decomposition based on Discrete Contour Evolution", Computer Vision and Image Understanding, vol. 73, pp. 441--454, 1999.

....video frames. Thus, a necessary condition for a simplification of a video polyline is that a sequence of vertices of the simplified polyline be a subsequence of the original one. Our approach to simplification of video polylines is based on a novel process of discrete curve evolution presented in [9] and applied in the context of shape similarity of planar objects in [11] However, here we will use a different measure of the relevance of vertices. Aside from its simplicity, the process of discrete curve evolution differs from the standard methods of polygonal approximation, like least square ....

Latecki, L.J., and Lakamper, R., "Convexity Rule for Shape Decomposition based on Discrete Contour Evolution", Computer Vision and Image Understanding, vol. 73, pp. 441--454, 1999.

No context found.

L. J. Latecki and R. Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. CVIU, 73(3):441--454, 1999.

No context found.

Latecki, L.J. & Lak amper, R. (1999). Convexity rule for shape decomposition based on discrete contour evolution. Computer Vision and Image Understanding 73:441--454.

No context found.

L. Latecki and R. Lakamper. Convexity rule for shape decomposition based on discrete contour evolution. CVIU, 73:441-- 454, 1999.

No context found.

L. J. Latecki and R. Lakmper, "Convexity Rule for Shape Decomposition Based on Discrete Contour Evolution," Computer Vision and Image Understanding (CVIU), Vol. 73, No. 3, pp. 441-454, 1999.

No context found.

L. J. Latecki, R. Lak amper. Convexity Rule for Shape Decomposition Based on Discrete Contour Evolution. Comput. Vision Image Understanding, 73 (3): 441-454, 1999.

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