Topology-Adaptive Modeling of Objects Using Surface Evolutions Based on 3D Mathematical Morphology

Size: px
Start display at page:

Download "Topology-Adaptive Modeling of Objects Using Surface Evolutions Based on 3D Mathematical Morphology"

Transcription

1 Systems and Computers in Japan, Vol. 33, No. 9, 2002 Translated from Denshi Joho Tsushin Gakkai Ronbunshi, Vol. J84-D-II, No. 5, May 2001, pp Topology-Adaptive Modeling of Objects Using Surface Evolutions Based on 3D Mathematical Morphology Kenji Hara, 1 Hongbin Zha, 2 and Tsutomu Hasegawa 3 1 Mechanics & Electronics Research Institute, Fukuoka Industrial Technology Center, Kitakyushu, Japan 2 Center for Information Science, Peking University, Beijing, China 3 Graduate School of Information Science and Electrical Engineering, Kyushu University, Fukuoka, Japan SUMMARY 1. Introduction Level set methods were proposed mainly by mathematicians for constructing a model of a 3D object of arbitrary topology. However, those methods are computationally inefficient due to repeated distance transformations and increased dimensions. In the paper, we propose a new method of modeling fast objects of arbitrary topology by using a surface evolution approach based on mathematical morphology. Given sensor data covering the whole object surface, the method begins with an initial approximation of the object by evolving a closed surface into a model topologically equivalent to the real object. The refined approximation is then performed using energy minimization. The method has been applied in several experiments using range data, and the results are reported in the paper Wiley Periodicals, Inc. Syst Comp Jpn, 33(9): 58 67, 2002; Published online in Wiley InterScience ( DOI /scj.1152 Key words: topology-adaptive object modeling; mathematical morphology; level set method; surface evolution; pseudo curvature flow. Recently, many methods have been proposed for automatic modeling of 3D free-form objects from multiview range images [1]. The resulting models, however, are usually still large data sets comprising great amounts of unorganized 3D points. To build a model base from which it is easy to extract the significant features such as surface normals or orientations, we have to transform the data further into surface models describing explicitly how those vertices are connected. The framework of optimization based on deformable surfaces was successfully used in the past for such object modeling [2 4]. Deformable surfaces have the ability to generate smooth closed surfaces on the basis of the principle of energy minimization. However, deformable surfaces have a serious drawback in modeling multiobject scenes or objects of arbitrary topology such as those with holes. One method for overcoming this problem is to apply shape evolution with blending [5]. However, while this method can allow the addition of a hole to the model, it cannot handle multiobject scenes adequately. On the other hand, topology-adaptive modeling method using a level set approach has been presented [6]. This level set modeling method can model multiobject scenes or objects of arbitrary topology by embedding an evolving surface (curve) as the zero level set of a function with a higher dimension [7, 8] Wiley Periodicals, Inc.

2 However, this modeling method is computationally inefficient since it requires repeated distance transformations and increased dimensions. In the paper, we propose a new fast method of topology-adaptive modeling by a surface evolution approach mainly based on 3D mathematical morphology. Mathematical morphology usually treats binary or gray scale images, and operates on them by morphological transformations using structuring elements [9]. This paper regards mathematical morphology as the dynamic model governing the 3D surface evolution, and then evolves surfaces by repeated morphology operations. It is well known that mathematical-morphology transformation based on a single dilation of a shape with a convex structuring element induces a differential deformation [10]. Thus, by introducing the curvature-dependent structuring elements, we can attain approximately but fast the surface deformations introduced in the level set approach, that is, curvature-dependent flow. First, by dilating every data point with a ball structuring element, we build a connected region, and then define it as the stopping region in which any surface point attains zero speed. Moreover, we set a volume including the whole 3D data, and then define the set of voxels lying on its boundary as the initial closed surface. In this paper, the surfaces are mainly represented by sets of the voxels interconnecting compactly. Next, we perform erosion transformation of the volume with a ball structuring element, and then extract the boundary of the transformed volume. This shrinking and splitting process is repeated until no surface point moves in the stopping region. Here, to keep the smoothness of the surface during its evolution, we use a structuring element with curvature-dependent size. After that, we extract the external boundary of the stopping region on the basis of the resulting surface, and then perform erosion along this boundary with the structuring element used in building the stopping region. By this updating of the stopping region, we shrink the surface further. We build a surface model based on mesh representation by viewing the surface as a set of hexahedra and then extracting those external squares. After the above rough estimation, by solving the optimal positions of those vertices on the basis of the principle of energy minimization, the resulting model is refined. The method has been applied in several experiments using synthetic and real images, and the results are reported in the paper. F. Here F can be a function of variables such as curvatures and normal directions [8]. The main idea of the level set method is to embed this moving surface Γ(t) as the zero level set of a higher-dimensional function u = φ(x, y, z, t). Let φ(x, t = 0), where x = (x, y, z), be defined by where d is the distance from x to Γ(t = 0), and the plus (minus) sign is chosen if the point x is outside (inside) Γ(t = 0). The stipulation that the zero level set of the function φ must always match the moving surface means that By the chain rule, we get The above equation is called the level set equation. One of the most desirable aspects of the level set formulation is that it allows the tracking of surfaces even as they undergo topological changes during evolution, as shown in Fig. 1. Malladi and colleagues proposed a method of extracting multiple contours from 2D images using the level set scheme [6]. They utilized a curve evolving at a curvaturedependent speed where ε is a positive constant and κ is the curvature. The problem here is how to define a speed function from the image data that serve as a halting criterion for this (1) (2) (3) (4) 2. Level Set Methods Given an initial 3D closed surface Γ(t = 0), we now derive a level-set formulation of the motion equations of the surface Γ(t) moving along its normal directions with speed Fig. 1. Curve evolution in a level set method. 59

3 speed function. Thus, by multiplying the above speed function by the filter we obtain the new image-based speed where G σ I denotes the image convolved with a Gaussian smoothing filter whose characteristic width is σ. The term G σ I is essentially zero except where the image gradient changes rapidly, in which case the value becomes large. The filter k I (x, y) is close to unity away from boundaries, and drops to zero near sharp changes in the image gradient, namely, in the stopping region. Thus, k I (x, y) serves to prevent the moving curve from passing out of the regions within which the original curves have to be constrained. To extend the above 2D modeling approach stated earlier to 3D modeling, one defines the level set equation as u = φ(x, y, z, t). However, this level set 3D modeling method has some difficulties, as follows. To determine the speed function F in the original method, searching for the closest surface points over the entire computational domain requires O(N 5 ) operations in 3D space, if we assume that there are N discretizing cells in each coordinate direction, plus approximately O(N 2 ) points on the surfaces. As an alternative, one can also perform searching only in a neighborhood of the zero level set; this is known as a narrow-band algorithm [12]. In this case, the operation count in 3D space drops to O(kN 2 ), where k is the number of cells in the narrow band. One of the other fast level set schemes is to apply an adaptive mesh in which each voxel has a size dependent on the curvature [13]. So far as we know, the fastest level set method is the fast marching method, which is a numerical technique for solving the eikonal equation (7). The computational cost is O(n 2 log n), where n is the number of points in each coordinate direction. However, this method is not computationally efficient yet. Thus, in this paper, we present a new method of modeling objects of arbitrary topology by using a morphology-based surface evolution. (5) (6) 3.1. Mathematical morphology and surface evolution Kimia and colleagues have shown that surface evolutions generated by 3D mathematical morphology operations are equivalent to those governed by a partial differential equation [10, 11]. For a ball structuring element, B, the set nb, where n is a positive integer, is defined as where 8 represents the dilation operator of 3D mathematical morphology. Here, if B is convex, nb is the same shape as B and n times as large as B. A single dilation with a structuring element nb is equivalent to n repeated dilations with B, which uses the distribution law in the dilation operation. Equation (8) shows that for a single operation with a convex structuring element, this operation can be decomposed as a sequence of n operations with the same structuring element but of size equal to 1/n of the original. In the limit as n, this morphology-based deformation can be viewed as a differential equation. On the other hand, the morphological transformation from a region S of 3D space to S can be regarded as deformation from its boundary S of S to S. In this paper, these boundaries are approximately described by sets of connected voxels. In object modeling, however, these sets of the points must be transformed to surface models by postprocessing. Now, let a shape S evolve by dilation (erosion) with a convex structuring element of size λ. Here, we assume that a point p on the surface S moves a distance Γ(p, t, λ) along the outward normal. Then, since Γ(p, t, 0) = 0, the speed F in the outward normal direction is represented as (7) (8) 3. Morphology-Based Surface Evolution To apply the surface evolution approach to topologyadaptive object modeling, we require that the surface retain smoothness and avoid self-intersection during its evolution. Though the level set approach satisfies these conditions, this method is computationally inefficient due to the distance transformations and the increased dimensions. In the section, we give a brief overview of the fast surface evolution approach in fitting closed surfaces to 3D data. It has been shown that dilation (erosion) operations can be regarded as surface evolutions governed by a partial differential equation expressed by Eq. (9). It is well known that the amount β(p, t) of differential deformation is the maximal projection of B onto the normal of S at p [10]. For (9) 60

4 example, for a unit ball structuring element B, we have the evolution equation (9) in the form This is the equation of a surface evolving at a constant speed along the normal. Differential equation (9) has been derived on the assumption that the surface S is smooth at point p; otherwise, even if p is a discontinuity, we can obtain the solution of Eq. (9) by using the level set method. It is known that this solution is equal to the above surface deformation by dilations (or erosions) [11] Pseudo curvature flow (10) Fig. 3. Amount of surface convexity. (12) The surface driven by Eq. (10) easily develops singularities during its evolution [Fig. 2(a)], and it cannot necessarily be applied to object modeling. In this subsection, we define the surface evolving while preserving smoothness and developing no self-intersections as a solution satisfying the equation in which Eq. (4) is extended to three dimensions. Differing from the level set approach [6, 8], this evolving surface can be traced at small computational cost without increased dimensions and repeated search for the closest points. In the paper, we let the closed surface smoothly contract and split by repeated erosions with a variable-size ball structuring element, and then apply this surface evolution approach to object modeling. The radius λ of the ball is defined as (11) where max (, 0) is a representation introduced to prevent λ from being negative. λ 0 and ε are positive integral and positive real constants, respectively. K = K(p, t) is the convexity of the surface, introduced instead of noise-sensitive 3D curvatures, and therefore, the second term of Eq. (11) can be regarded as a smoothing term. K is defined as the volume rate of the region outside the surface in the neighborhood of p, and thus it is defined as where α = α(p, t) and β = β(p, t) are the numbers of inside and outside cells, respectively, within a distance ρ from p (Fig. 3). In the above surface evolution process, the surface smoothness is preserved as the mean-curvature flow defined in Eq. (4) [Fig. 2(b)], and therefore, we call it pseudo curvature flow. When dilations are used, Eq. (11) is replaced by λ = max (λ 0 ε K, 0). Here, we discuss pseudo curvature flow from the viewpoint of computational expense. There are approximately O(n 2 ) voxels in the evolving surface, assuming that there are n voxels in each coordinate direction. First, to determine the size λ of the structuring element for these voxels, the approximate value K of the mean curvature must be calculated. However, it only takes values equal to the weighted sum of neighboring voxels. Next, we must calculate the region to be eroded for each voxel in the surface. We prepare a set of several ball structuring elements of different sizes as (13) where B is the ball structuring element of unit size and λ 0 is the positive integer in Eq. (11). After that, we take the structuring element from G which has the size of the integral part of λ obtained by Eq. (11). If the integral part of λ is larger than 2λ 0, the structuring element of the size of 2λ 0 is selected. Consequently, it is obvious that pseudo curvature flow results in a time complexity of O(n 2 ). 4. Topology-Adaptive Modeling by Surface Evolution Fig. 2. Surface evolution using size-variable structuring element. One of the difficulties arising in modeling 3D objects from range data by the surface evolution approach is determining how to halt the evolving surface at the object surface. In our strategy, we define the region in which any surface point halts moving, and therefore, a rough but topologically correct model can be obtained by surface 61

5 evolution based on pseudo curvature flow; we then refine this model by energy minimization. In this section, we give an overview of this topology-adaptive object modeling method based on both surface evolution and the conventional deformable surface Topology-adaptive modeling algorithm The algorithm we developed consists of a two-phase procedure. First, by dilating a given set of 3D image data [Fig. 4(a)] with the ball structuring element of size δ, we build a connected region [Fig. 4(b)]. This region is called the stopping region, in which any surface point halts moving. Next, we set a single closed surface surrounding the whole body of data as the initial shape, and then repeat the processes of shrinking and splitting by the pseudo curvature flow until every surface point stops moving near the border of the stopping region [Fig. 4(c)]. After that, we separate the surface of the stopping region into the boundary surface lying inside the object (internal boundary surface) and the other surface (external boundary surface) on the basis of the resulting surface. We also erode the stopping region along the external boundary surface with a ball structuring element of size δ, which updates the stopping region [Fig. 4(d)]. Within this stopping region, we shrink the surface again until it does not evolve further [Fig. 4(e)]. We build a surface model based on a mesh representation by viewing the surface as a set of hexahedra and then extracting those external squares. The above step for initial estimation corresponds to the first stage. At the second stage, the surface model resulting from the above step is refined for the purpose of improving its accuracy. The optimal positions of those vertices are calculated based on the energy minimization principle [Fig. 4(f)]. Further details are described below Initial estimation Given a set of 3D image data, our goal here is to estimate a topologically correct model by surface evolution based on pseudo curvature flow. We expand every data point into a ball with radius δ by eroding it with a ball structuring element δb of size δ. If δ is sufficiently large, the expanded data points are connected with each other. However, inappropriate selection of the δ values causes a failure in modeling due to the topologically wrong stopping regions. The stopping regions resulting from too large δ and small δ are illustrated in Figs. 5(c) and 5(d), respectively. It is seen that both are topologically wrong. We determine the δ value as follows. We first build a stopping region with comparatively large δ, and then extract the voxels of its boundary. After that, we repeat the process consisting of decreasing δ and extracting the border surface of the stopping region until the stopping region is regarded as being topologically equivalent to the object surface, and thus obtain the estimated value of δ [Fig. 5(b)]. Here, as a criterion for the topology judgment, we introduce the detection rate (14) where N r is the number of data points having model points within a distance δ, and N is the total number of data points. Fig. 4. Outline of our algorithm. 62

6 Fig. 6. Surface shrinking-and-splitting. Fig. 5. Influence of variation of δ. Here, if q is smaller than a threshold th, δ is lowered, since it is inferred that the current model is wrong in its topology. From the above step, a set of discrete data is transformed into the stopping region X, which is between the external border surface and the internal one, and X is represented as (15) where L is a set of data points. Next, we set a single closed surface surrounding the whole body of data as the initial shape. Then, we shrink and split this surface by repeated erosions with the variable-size ball structuring element λb until every surface point stops moving (Fig. 6). Here, the radius λ = λ(p, t) of the ball structuring element centered on a point p of the evolving surface at time t is defined by modifying Eq. (11) as with this corrected stopping region, we can obtain a surface stopping near the surface of the object. However, the resulting model is still a set of voxels, and thus, it is not a surface model explicitly describing how the vertices are connected. To transform the set of voxels into a polygonal mesh, we regard the surface as a set of hexahedra, and then extract their external squares. In addition, we can decrease the number of patches by dividing the 3D space into larger sections Model refinement This stage is employed mainly to enhance the accuracy of modeling by removing local irregularities resulting from the earlier evolution. The current surface model provides a reasonable approximation to the object surface in the topology. Moreover, it has a comparatively smooth shape owing to the smoothing effect of pseudo curvature flow. Therefore, we can apply conventional deformable surface techniques to this model [2 4]. For the set v of position vectors of the vertices, the energy function to be minimized can be expressed by (16) (17) where λ 0 and ε are the same positive constants as those used in Eq. (11). K = K(p, t) is the amount of convexity of the surface defined in Eq. (12). Equation (16) states that p is not moved if λ 0 + ε K is negative or p lies in X, and thus, the surface stops evolving so that the object is covered by the surface [Fig. 6(d)]. After that, we extract the external boundary surface of the stopping region X by checking whether each voxel of the boundary surface of the stopping region lies outside this surface or not. Moreover, we erode X along this external boundary surface with the structuring element δb used earlier. By restarting surface evolution E int (v) corresponds to a smoothness constraint based on springs imaginarily placed between neighboring vertices. E est (v) is a data closeness term based on springs placed between the vertices and their nearest data points [3]. Equation (17) can be minimized fast using a greedy algorithm [14]. As a result, we can obtain the optimum mesh fitted to 3D data. 5. Experimental Results In this section, we present the results of some experiments using synthetic range data and real data acquired by a laser range finder. 63

7 Figure 7(a) shows 3D data synthesized from CAD data for a hexagon nut. This experiment is made mainly to show the capability of our method for modeling sparse data and its superiority to the conventional level set modeling method. At the beginning, we specify a cubic closed surface covering the data, and select δ as 8 using the parameter estimation technique in Section 4.2. Next, we shrink and split the surface until no surface point moves as shown in Fig. 7(b). After that, we update the stopping region, and obtain the model for this region shown in Fig. 7(c). We obtain the final model by optimizing the positions of vertices in the model refinement stage in Fig. 7(d). Since the object has a simple shape, we reduce the number of patches to about 1/4 in extracting the mesh from the object model shown in Fig. 7(c) by dividing the 3D space into double sections. To compare our method with the existing topologyadaptive modeling method, intermediate results obtained using the fast level set approach called the narrow-band algorithm [12] are shown in Figs. 8(a) and 8(b). These models correspond to Figs. 7(b) and 7(c), respectively. As is evident here, there is not much difference between them. Table 1 summarizes the surface evolution process in the first row, indicating the values in Figs. 7 and 8. The subsequent rows correspond to the widths of stopping region (δ), numbers of vertices (N), and used CPU times (T) in seconds on a Pentium II 300-MHz personal computer. Comparing the two CPU times corresponding to Figs. 7(c) and 8(b) in Table 1, our method is approximately Fig. 8. Experiment by a level set modeling method. 10 times faster than level set modeling method using the narrow-band technique. Figure 9 illustrates the process of surface evolution in the initial estimation stage by cross-sectional views. Figures 9(b) and 9(c) illustrate the intermediate and final results, respectively, obtained by surface shrinking and splitting for the initial stopping region. Figures 10(a) to 10(c) show cross sections of the evolving surface when no stopping region is set. The smoothness of the surface is gradually increased in each step due to the curvature-dependent speed. In the second experiment, the object to be modeled is a teacup with a hole and rim shown in Fig. 11(a). This object is a typical difficult-to-model example. If we connect the data points in the neighborhood one by one, a data point on the inside of this cup may be connected with one on the outside. While the deformable surfaces can avoid this problem, the problem of how to restore the hole formed by the handle arises. We obtained images of a teacup suspended by a wire from several viewpoints. Figure 11(b) shows the 3D data of this object. The intermediate and final results are illustrated in Figs. 12(a) and 12(b). The number of vertices is 24,352 and the CPU time is 165 seconds on a Pentium III 600-MHz personal computer. The final experiment aims to show the capability of our method in modeling multiobject scenes completely without a priori knowledge about the topology. The objects to be modeled are three statues, including one with a hole, shown in Fig. 13(a). Figure 13(b) shows the 3D data. Table 1. Result summary of the experiment in Section 5.1 Fig. 7. Experiment for an object with a hole. 64

8 Fig. 9. Process of topological changes. Fig. 13. Multiple objects. Fig. 10. Process of surface evolving with smoothness. Figures 14(a) to 14(d) show the intermediate results. Here, the single closed surface shrinks as shown in Figs. 14(a) to Fig. 14(c), and then splits into three surfaces while developing a hole as shown in Fig. 14(d). The results at the first stage and the final stage are shown in Figs. 14(e) and 14(f), respectively. The number of vertices is 21,144 and the CPU time is 2276 seconds on the same computer as in the first experiment. Fig. 11. Object with a hole and a rim. Fig. 12. Experiment for an object with a hole and a rim. Fig. 14. Experiment for multiple objects. 65

9 6. Conclusions In the paper, we have proposed a method of modeling a 3D object of arbitrary topology by using variable-size ball marching. Our algorithm has the advantage of modeling objects fast and easily. It can also be implemented without any knowledge about correspondences between data points and sensor viewpoints, which constitutes the main difference between our method and the statistical level set method [15]. In the future, we will investigate more deeply the problem of designing the criterion of topological consistency between the model and 3D data in building the stopping region. In addition, we must analyze the proposed marching method with greater mathematical rigor. REFERENCES 1. Zha H, Morooka K, Hasegawa T. Active modeling of 3D objects: Planning on the next best pose (NBP) for acquiring range images. Proc Int Conf on Recent Advances in 3-D Digital Imaging and Modeling, p 68 75, Delingette H, Hebert M, Ikeuchi K. Shape representation and image segmentation using deformable surfaces. Proc Conf on Computer Vision and Pattern Recognition, p , McInerney T, Terzopoulos D. A finite element model for 3D shape reconstruction and nonrigid motion tracking. Proc Int Conf on Computer Vision, p , Hara K, Zha H, Hasegawa T. Regularization based 3-D object modeling from multiple range images. Proc Int Conf on Pattern Recognition 1998;2: DeCarlo D, Metaxas D. Shape evolution with structural and topological changes using blending. IEEE Trans Pattern Anal Machine Intell 1998;20: Malladi R, Sethian JA, Vemuri BC. Shape modeling with front propagation: A level set approach. IEEE Trans Pattern Anal Machine Intell 1995;17: Sethian JA. Level set methods and fast marching methods. Cambridge University Press; Osher S, Sethian JA. Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton Jacobi formulation. J Comput Phys 1988;79: Serra J. Image analysis and mathematical morphology. Academic Press; Arehart AB, Vincent L, Kimia BB. Mathematical morphology: The Hamilton Jacobi connection. Proc Int Conf on Computer Vision, p , Kimia BB, Tannenbaum AR, Zucker SW. Shapes, shocks, and deformations, I: The components of shape and the reaction-diffusion space. Int J Comput Vision 1995;15: Chopp DL. Computing minimal surfaces via level set curvature flow. J Comput Phys 1993;106: Milne B. Adaptive level set methods interfaces. J Comput Phys, Ph.D. thesis, Department of Mathematics, University of California, Berkeley, Williams DJ, Shah M. A fast algorithm for active contours. Proc Int Conf on Computer Vision, p , Whitaker RT. A level-set approach to 3D reconstruction from range data. Int J Comput Vision 1998;29: AUTHORS Kenji Hara received his M.S. degree in mathematics and sciences from Kyoto University in 1984 and Ph.D. degree from Kyushu University in He joined Takeda Chemical Industries Co., Ltd. in 1989, where he engaged in software development as a system engineer. From 1991 to 2000, he was associated with the Mechanics & Electronics Research Institute at Fukuoka Industrial Technology Center. He is currently an adjunctive researcher at the Institute of Industrial Science, University of Tokyo. His current research interests include physics-based vision, 3D object recognition, and automatic geometric modeling. 66

10 AUTHORS (continued) (from left to right) Hongbin Zha received his B.E. degree in electrical engineering from Hefei University of Technology, China, in 1983, and M.S. and Ph.D. degrees in electrical engineering from Kyushu University in 1987 and From 1990 to 1991, he was a research associate in the Department of Control Engineering and Science at Kyushu Institute of Technology. From 1991 to 2000, he was with the Department of Computer Science and Communication Engineering, and then the Department of Intelligent Systems, Graduate School of Information Science and Electrical Engineering, Kyushu University, as an associate professor. He is currently a professor at the Center for Information Science, Peking University, China. His research interests include robotics, computer vision, automatic geometric modeling, and sensor-based interaction between robots or robots and environments. Tsutomu Hasegawa received his B.E. degree in electronic engineering and Ph.D. degree from Tokyo Institute of Technology in 1973 and He was associated with the Electrotechnical Laboratory of the Japanese Government from 1973 to 1992, where he performed research in robotics. From 1981 to 1982, he was a visiting researcher at the Laboratoire d Automatique et d Analyses des Systemes, Toulouse, France. He joined Kyushu University in 1992, and is currently a professor in the Deparment of Intelligent Systems, Graduate School of Information Science and Electrical Engineering. His research interests are in manipulator control, geometric modeling and reasoning, motion planning, and man machine interaction. 67

Automated Segmentation Using a Fast Implementation of the Chan-Vese Models

Automated Segmentation Using a Fast Implementation of the Chan-Vese Models Automated Segmentation Using a Fast Implementation of the Chan-Vese Models Huan Xu, and Xiao-Feng Wang,,3 Intelligent Computation Lab, Hefei Institute of Intelligent Machines, Chinese Academy of Science,

More information

Dr. Ulas Bagci

Dr. Ulas Bagci Lecture 9: Deformable Models and Segmentation CAP-Computer Vision Lecture 9-Deformable Models and Segmentation Dr. Ulas Bagci bagci@ucf.edu Lecture 9: Deformable Models and Segmentation Motivation A limitation

More information

The Level Set Method. Lecture Notes, MIT J / 2.097J / 6.339J Numerical Methods for Partial Differential Equations

The Level Set Method. Lecture Notes, MIT J / 2.097J / 6.339J Numerical Methods for Partial Differential Equations The Level Set Method Lecture Notes, MIT 16.920J / 2.097J / 6.339J Numerical Methods for Partial Differential Equations Per-Olof Persson persson@mit.edu March 7, 2005 1 Evolving Curves and Surfaces Evolving

More information

HIGH DENSITY PLASMA DEPOSITION MODELING USING LEVEL SET METHODS

HIGH DENSITY PLASMA DEPOSITION MODELING USING LEVEL SET METHODS HIGH DENSITY PLASMA DEPOSITION MODELING USING LEVEL SET METHODS D. Adalsteinsson J.A. Sethian Dept. of Mathematics University of California, Berkeley 94720 and Juan C. Rey Technology Modeling Associates

More information

Filters. Advanced and Special Topics: Filters. Filters

Filters. Advanced and Special Topics: Filters. Filters Filters Advanced and Special Topics: Filters Dr. Edmund Lam Department of Electrical and Electronic Engineering The University of Hong Kong ELEC4245: Digital Image Processing (Second Semester, 2016 17)

More information

Medical Image Segmentation using Level Sets

Medical Image Segmentation using Level Sets Medical Image Segmentation using Level Sets Technical Report #CS-8-1 Tenn Francis Chen Abstract Segmentation is a vital aspect of medical imaging. It aids in the visualization of medical data and diagnostics

More information

Edge and local feature detection - 2. Importance of edge detection in computer vision

Edge and local feature detection - 2. Importance of edge detection in computer vision Edge and local feature detection Gradient based edge detection Edge detection by function fitting Second derivative edge detectors Edge linking and the construction of the chain graph Edge and local feature

More information

Extract Object Boundaries in Noisy Images using Level Set. Literature Survey

Extract Object Boundaries in Noisy Images using Level Set. Literature Survey Extract Object Boundaries in Noisy Images using Level Set by: Quming Zhou Literature Survey Submitted to Professor Brian Evans EE381K Multidimensional Digital Signal Processing March 15, 003 Abstract Finding

More information

Level Set Models for Computer Graphics

Level Set Models for Computer Graphics Level Set Models for Computer Graphics David E. Breen Department of Computer Science Drexel University Ross T. Whitaker School of Computing University of Utah Ken Museth Department of Science and Technology

More information

A Singular Example for the Averaged Mean Curvature Flow

A Singular Example for the Averaged Mean Curvature Flow To appear in Experimental Mathematics Preprint Vol. No. () pp. 3 7 February 9, A Singular Example for the Averaged Mean Curvature Flow Uwe F. Mayer Abstract An embedded curve is presented which under numerical

More information

Isophote-Based Interpolation

Isophote-Based Interpolation Isophote-Based Interpolation Bryan S. Morse and Duane Schwartzwald Department of Computer Science, Brigham Young University 3361 TMCB, Provo, UT 84602 {morse,duane}@cs.byu.edu Abstract Standard methods

More information

Image Segmentation Techniques for Object-Based Coding

Image Segmentation Techniques for Object-Based Coding Image Techniques for Object-Based Coding Junaid Ahmed, Joseph Bosworth, and Scott T. Acton The Oklahoma Imaging Laboratory School of Electrical and Computer Engineering Oklahoma State University {ajunaid,bosworj,sacton}@okstate.edu

More information

Color Image Segmentation Editor Based on the Integration of Edge-Linking, Region Labeling and Deformable Model

Color Image Segmentation Editor Based on the Integration of Edge-Linking, Region Labeling and Deformable Model This paper appears in: IEEE International Conference on Systems, Man and Cybernetics, 1999 Color Image Segmentation Editor Based on the Integration of Edge-Linking, Region Labeling and Deformable Model

More information

Topic 6 Representation and Description

Topic 6 Representation and Description Topic 6 Representation and Description Background Segmentation divides the image into regions Each region should be represented and described in a form suitable for further processing/decision-making Representation

More information

Binary Image Processing. Introduction to Computer Vision CSE 152 Lecture 5

Binary Image Processing. Introduction to Computer Vision CSE 152 Lecture 5 Binary Image Processing CSE 152 Lecture 5 Announcements Homework 2 is due Apr 25, 11:59 PM Reading: Szeliski, Chapter 3 Image processing, Section 3.3 More neighborhood operators Binary System Summary 1.

More information

Binary Shape Characterization using Morphological Boundary Class Distribution Functions

Binary Shape Characterization using Morphological Boundary Class Distribution Functions Binary Shape Characterization using Morphological Boundary Class Distribution Functions Marcin Iwanowski Institute of Control and Industrial Electronics, Warsaw University of Technology, ul.koszykowa 75,

More information

C E N T E R A T H O U S T O N S C H O O L of H E A L T H I N F O R M A T I O N S C I E N C E S. Image Operations II

C E N T E R A T H O U S T O N S C H O O L of H E A L T H I N F O R M A T I O N S C I E N C E S. Image Operations II T H E U N I V E R S I T Y of T E X A S H E A L T H S C I E N C E C E N T E R A T H O U S T O N S C H O O L of H E A L T H I N F O R M A T I O N S C I E N C E S Image Operations II For students of HI 5323

More information

Unstructured Mesh Generation for Implicit Moving Geometries and Level Set Applications

Unstructured Mesh Generation for Implicit Moving Geometries and Level Set Applications Unstructured Mesh Generation for Implicit Moving Geometries and Level Set Applications Per-Olof Persson (persson@mit.edu) Department of Mathematics Massachusetts Institute of Technology http://www.mit.edu/

More information

A Geometric Contour Framework with Vector Field Support

A Geometric Contour Framework with Vector Field Support A Geometric Contour Framework with Vector Field Support Zhenglong Li, Qingshan Liu, and Hanqing Lu National Laboratory of Pattern Recognition Automation of Institute, Chinese Academy of Sciences P.O. Box

More information

IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 17, NO. 5, MAY

IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 17, NO. 5, MAY IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 17, NO. 5, MAY 2008 645 A Real-Time Algorithm for the Approximation of Level-Set-Based Curve Evolution Yonggang Shi, Member, IEEE, and William Clem Karl, Senior

More information

Morphological Image Processing

Morphological Image Processing Morphological Image Processing Morphology Identification, analysis, and description of the structure of the smallest unit of words Theory and technique for the analysis and processing of geometric structures

More information

Recognizing Buildings in Urban Scene of Distant View ABSTRACT

Recognizing Buildings in Urban Scene of Distant View ABSTRACT Recognizing Buildings in Urban Scene of Distant View Peilin Liu, Katsushi Ikeuchi and Masao Sakauchi Institute of Industrial Science, University of Tokyo, Japan 7-22-1 Roppongi, Minato-ku, Tokyo 106, Japan

More information

An Active Contour Model without Edges

An Active Contour Model without Edges An Active Contour Model without Edges Tony Chan and Luminita Vese Department of Mathematics, University of California, Los Angeles, 520 Portola Plaza, Los Angeles, CA 90095-1555 chan,lvese@math.ucla.edu

More information

CHAPTER 6 DETECTION OF MASS USING NOVEL SEGMENTATION, GLCM AND NEURAL NETWORKS

CHAPTER 6 DETECTION OF MASS USING NOVEL SEGMENTATION, GLCM AND NEURAL NETWORKS 130 CHAPTER 6 DETECTION OF MASS USING NOVEL SEGMENTATION, GLCM AND NEURAL NETWORKS A mass is defined as a space-occupying lesion seen in more than one projection and it is described by its shapes and margin

More information

Segmentation and Tracking of Partial Planar Templates

Segmentation and Tracking of Partial Planar Templates Segmentation and Tracking of Partial Planar Templates Abdelsalam Masoud William Hoff Colorado School of Mines Colorado School of Mines Golden, CO 800 Golden, CO 800 amasoud@mines.edu whoff@mines.edu Abstract

More information

Biomedical Image Analysis. Mathematical Morphology

Biomedical Image Analysis. Mathematical Morphology Biomedical Image Analysis Mathematical Morphology Contents: Foundation of Mathematical Morphology Structuring Elements Applications BMIA 15 V. Roth & P. Cattin 265 Foundations of Mathematical Morphology

More information

Multi-Scale Free-Form Surface Description

Multi-Scale Free-Form Surface Description Multi-Scale Free-Form Surface Description Farzin Mokhtarian, Nasser Khalili and Peter Yuen Centre for Vision Speech and Signal Processing Dept. of Electronic and Electrical Engineering University of Surrey,

More information

A FAST IMPLEMENTATION OF THE LEVEL SET METHOD WITHOUT SOLVING PARTIAL DIFFERENTIAL EQUATIONS. Yonggang Shi, William Clem Karl

A FAST IMPLEMENTATION OF THE LEVEL SET METHOD WITHOUT SOLVING PARTIAL DIFFERENTIAL EQUATIONS. Yonggang Shi, William Clem Karl A FAST IMPLEMENTATION OF THE LEVEL SET METHOD WITHOUT SOLVING PARTIAL DIFFERENTIAL EQUATIONS Yonggang Shi, William Clem Karl January, 2005 Boston University Department of Electrical and Computer Engineering

More information

Morphological Image Processing

Morphological Image Processing Morphological Image Processing Binary image processing In binary images, we conventionally take background as black (0) and foreground objects as white (1 or 255) Morphology Figure 4.1 objects on a conveyor

More information

I. INTRODUCTION. Figure-1 Basic block of text analysis

I. INTRODUCTION. Figure-1 Basic block of text analysis ISSN: 2349-7637 (Online) (RHIMRJ) Research Paper Available online at: www.rhimrj.com Detection and Localization of Texts from Natural Scene Images: A Hybrid Approach Priyanka Muchhadiya Post Graduate Fellow,

More information

DEVELOPMENT OF POSITION MEASUREMENT SYSTEM FOR CONSTRUCTION PILE USING LASER RANGE FINDER

DEVELOPMENT OF POSITION MEASUREMENT SYSTEM FOR CONSTRUCTION PILE USING LASER RANGE FINDER S17- DEVELOPMENT OF POSITION MEASUREMENT SYSTEM FOR CONSTRUCTION PILE USING LASER RANGE FINDER Fumihiro Inoue 1 *, Takeshi Sasaki, Xiangqi Huang 3, and Hideki Hashimoto 4 1 Technica Research Institute,

More information

Lecture 12 Level Sets & Parametric Transforms. sec & ch. 11 of Machine Vision by Wesley E. Snyder & Hairong Qi

Lecture 12 Level Sets & Parametric Transforms. sec & ch. 11 of Machine Vision by Wesley E. Snyder & Hairong Qi Lecture 12 Level Sets & Parametric Transforms sec. 8.5.2 & ch. 11 of Machine Vision by Wesley E. Snyder & Hairong Qi Spring 2017 16-725 (CMU RI) : BioE 2630 (Pitt) Dr. John Galeotti The content of these

More information

Local Patterns and Connectivity Indexes in a Three Dimensional Digital Picture

Local Patterns and Connectivity Indexes in a Three Dimensional Digital Picture Review Forma, 17, 275 291, 2002 Local Patterns and Connectivity Indexes in a Three Dimensional Digital Picture Junichiro TORIWAKI 1 * and Tatsuhiro YONEKURA 2 1 Department of Information Engineering, Graduate

More information

2.2 Weighting function

2.2 Weighting function Annual Report (23) Kawahara Lab. On Shape Function of Element-Free Galerkin Method for Flow Analysis Daigo NAKAI and Mutsuto KAWAHARA Department of Civil Engineering, Chuo University, Kasuga 3 27, Bunkyo

More information

Image Segmentation. Ross Whitaker SCI Institute, School of Computing University of Utah

Image Segmentation. Ross Whitaker SCI Institute, School of Computing University of Utah Image Segmentation Ross Whitaker SCI Institute, School of Computing University of Utah What is Segmentation? Partitioning images/volumes into meaningful pieces Partitioning problem Labels Isolating a specific

More information

STATISTICAL ATLAS-BASED SUB-VOXEL SEGMENTATION OF 3D BRAIN MRI

STATISTICAL ATLAS-BASED SUB-VOXEL SEGMENTATION OF 3D BRAIN MRI STATISTICA ATAS-BASED SUB-VOXE SEGMENTATION OF 3D BRAIN MRI Marcel Bosc 1,2, Fabrice Heitz 1, Jean-Paul Armspach 2 (1) SIIT UMR-7005 CNRS / Strasbourg I University, 67400 Illkirch, France (2) IPB UMR-7004

More information

EECS490: Digital Image Processing. Lecture #17

EECS490: Digital Image Processing. Lecture #17 Lecture #17 Morphology & set operations on images Structuring elements Erosion and dilation Opening and closing Morphological image processing, boundary extraction, region filling Connectivity: convex

More information

Mathematical Morphology and Distance Transforms. Robin Strand

Mathematical Morphology and Distance Transforms. Robin Strand Mathematical Morphology and Distance Transforms Robin Strand robin.strand@it.uu.se Morphology Form and structure Mathematical framework used for: Pre-processing Noise filtering, shape simplification,...

More information

Isophote-Based Interpolation

Isophote-Based Interpolation Brigham Young University BYU ScholarsArchive All Faculty Publications 1998-10-01 Isophote-Based Interpolation Bryan S. Morse morse@byu.edu Duane Schwartzwald Follow this and additional works at: http://scholarsarchive.byu.edu/facpub

More information

Separation of Position and Direction Information of Robots by a Product Model of Self-Organizing Map and Neural Gas

Separation of Position and Direction Information of Robots by a Product Model of Self-Organizing Map and Neural Gas Systems and Computers in Japan, Vol. 36, No. 11, 2005 Translated from Denshi Joho Tsushin Gakkai Ronbunshi, Vol. J87-D-II, No. 7, July 2004, pp. 1529 1538 Separation of Position and Direction Information

More information

Module 1 Lecture Notes 2. Optimization Problem and Model Formulation

Module 1 Lecture Notes 2. Optimization Problem and Model Formulation Optimization Methods: Introduction and Basic concepts 1 Module 1 Lecture Notes 2 Optimization Problem and Model Formulation Introduction In the previous lecture we studied the evolution of optimization

More information

Level Set Methods and Fast Marching Methods

Level Set Methods and Fast Marching Methods Level Set Methods and Fast Marching Methods I.Lyulina Scientific Computing Group May, 2002 Overview Existing Techniques for Tracking Interfaces Basic Ideas of Level Set Method and Fast Marching Method

More information

Morphological Image Processing

Morphological Image Processing Morphological Image Processing Ranga Rodrigo October 9, 29 Outline Contents Preliminaries 2 Dilation and Erosion 3 2. Dilation.............................................. 3 2.2 Erosion..............................................

More information

Accurate and Dense Wide-Baseline Stereo Matching Using SW-POC

Accurate and Dense Wide-Baseline Stereo Matching Using SW-POC Accurate and Dense Wide-Baseline Stereo Matching Using SW-POC Shuji Sakai, Koichi Ito, Takafumi Aoki Graduate School of Information Sciences, Tohoku University, Sendai, 980 8579, Japan Email: sakai@aoki.ecei.tohoku.ac.jp

More information

Chapter 8 Visualization and Optimization

Chapter 8 Visualization and Optimization Chapter 8 Visualization and Optimization Recommended reference books: [1] Edited by R. S. Gallagher: Computer Visualization, Graphics Techniques for Scientific and Engineering Analysis by CRC, 1994 [2]

More information

Topologically Adaptable Snakes

Topologically Adaptable Snakes Published in the Proc. of the Fifth Int. Conf. on Computer Vision (ICCV 95), Cambridge, MA, USA, June, 1995, 840 845.840 Topologically Adaptable Snakes Tim McInerney and Demetri Terzopoulos Department

More information

Design of Low-Delay FIR Half-Band Filters with Arbitrary Flatness and Its Application to Filter Banks

Design of Low-Delay FIR Half-Band Filters with Arbitrary Flatness and Its Application to Filter Banks Electronics and Communications in Japan, Part 3, Vol 83, No 10, 2000 Translated from Denshi Joho Tsushin Gakkai Ronbunshi, Vol J82-A, No 10, October 1999, pp 1529 1537 Design of Low-Delay FIR Half-Band

More information

Digital Image Processing Fundamentals

Digital Image Processing Fundamentals Ioannis Pitas Digital Image Processing Fundamentals Chapter 7 Shape Description Answers to the Chapter Questions Thessaloniki 1998 Chapter 7: Shape description 7.1 Introduction 1. Why is invariance to

More information

weighted minimal surface model for surface reconstruction from scattered points, curves, and/or pieces of surfaces.

weighted minimal surface model for surface reconstruction from scattered points, curves, and/or pieces of surfaces. weighted minimal surface model for surface reconstruction from scattered points, curves, and/or pieces of surfaces. joint work with (S. Osher, R. Fedkiw and M. Kang) Desired properties for surface reconstruction:

More information

Level set methods Formulation of Interface Propagation Boundary Value PDE Initial Value PDE Motion in an externally generated velocity field

Level set methods Formulation of Interface Propagation Boundary Value PDE Initial Value PDE Motion in an externally generated velocity field Level Set Methods Overview Level set methods Formulation of Interface Propagation Boundary Value PDE Initial Value PDE Motion in an externally generated velocity field Convection Upwind ddifferencingi

More information

Fast Distance Transform Computation using Dual Scan Line Propagation

Fast Distance Transform Computation using Dual Scan Line Propagation Fast Distance Transform Computation using Dual Scan Line Propagation Fatih Porikli Tekin Kocak Mitsubishi Electric Research Laboratories, Cambridge, USA ABSTRACT We present two fast algorithms that approximate

More information

3D Shape Recovery of Smooth Surfaces: Dropping the Fixed Viewpoint Assumption

3D Shape Recovery of Smooth Surfaces: Dropping the Fixed Viewpoint Assumption IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL., NO., 1 3D Shape Recovery of Smooth Surfaces: Dropping the Fixed Viewpoint Assumption Yael Moses Member, IEEE and Ilan Shimshoni Member,

More information

College of Engineering, Trivandrum.

College of Engineering, Trivandrum. Analysis of CT Liver Images Using Level Sets with Bayesian Analysis-A Hybrid Approach Sajith A.G 1, Dr. Hariharan.S 2 1 Research Scholar, 2 Professor, Department of Electrical&Electronics Engineering College

More information

Occlusion Detection of Real Objects using Contour Based Stereo Matching

Occlusion Detection of Real Objects using Contour Based Stereo Matching Occlusion Detection of Real Objects using Contour Based Stereo Matching Kenichi Hayashi, Hirokazu Kato, Shogo Nishida Graduate School of Engineering Science, Osaka University,1-3 Machikaneyama-cho, Toyonaka,

More information

A MORPHOLOGY-BASED FILTER STRUCTURE FOR EDGE-ENHANCING SMOOTHING

A MORPHOLOGY-BASED FILTER STRUCTURE FOR EDGE-ENHANCING SMOOTHING Proceedings of the 1994 IEEE International Conference on Image Processing (ICIP-94), pp. 530-534. (Austin, Texas, 13-16 November 1994.) A MORPHOLOGY-BASED FILTER STRUCTURE FOR EDGE-ENHANCING SMOOTHING

More information

Segmentation Using Active Contour Model and Level Set Method Applied to Medical Images

Segmentation Using Active Contour Model and Level Set Method Applied to Medical Images Segmentation Using Active Contour Model and Level Set Method Applied to Medical Images Dr. K.Bikshalu R.Srikanth Assistant Professor, Dept. of ECE, KUCE&T, KU, Warangal, Telangana, India kalagaddaashu@gmail.com

More information

Level Set Method in a Finite Element Setting

Level Set Method in a Finite Element Setting Level Set Method in a Finite Element Setting John Shopple University of California, San Diego November 6, 2007 Outline 1 Level Set Method 2 Solute-Solvent Model 3 Reinitialization 4 Conclusion Types of

More information

Edge-Preserving Denoising for Segmentation in CT-Images

Edge-Preserving Denoising for Segmentation in CT-Images Edge-Preserving Denoising for Segmentation in CT-Images Eva Eibenberger, Anja Borsdorf, Andreas Wimmer, Joachim Hornegger Lehrstuhl für Mustererkennung, Friedrich-Alexander-Universität Erlangen-Nürnberg

More information

Numerical schemes for Hamilton-Jacobi equations, control problems and games

Numerical schemes for Hamilton-Jacobi equations, control problems and games Numerical schemes for Hamilton-Jacobi equations, control problems and games M. Falcone H. Zidani SADCO Spring School Applied and Numerical Optimal Control April 23-27, 2012, Paris Lecture 2/3 M. Falcone

More information

EE795: Computer Vision and Intelligent Systems

EE795: Computer Vision and Intelligent Systems EE795: Computer Vision and Intelligent Systems Spring 2012 TTh 17:30-18:45 WRI C225 Lecture 04 130131 http://www.ee.unlv.edu/~b1morris/ecg795/ 2 Outline Review Histogram Equalization Image Filtering Linear

More information

Face Hallucination Based on Eigentransformation Learning

Face Hallucination Based on Eigentransformation Learning Advanced Science and Technology etters, pp.32-37 http://dx.doi.org/10.14257/astl.2016. Face allucination Based on Eigentransformation earning Guohua Zou School of software, East China University of Technology,

More information

Implicit Surface Reconstruction from 3D Scattered Points Based on Variational Level Set Method

Implicit Surface Reconstruction from 3D Scattered Points Based on Variational Level Set Method Implicit Surface econstruction from D Scattered Points Based on Variational Level Set Method Hanbo Liu Department ofshenzhen graduate school, Harbin Institute oftechnology, Shenzhen, 58055, China liu_hanbo@hit.edu.cn

More information

A Systematic Analysis System for CT Liver Image Classification and Image Segmentation by Local Entropy Method

A Systematic Analysis System for CT Liver Image Classification and Image Segmentation by Local Entropy Method A Systematic Analysis System for CT Liver Image Classification and Image Segmentation by Local Entropy Method A.Anuja Merlyn 1, A.Anuba Merlyn 2 1 PG Scholar, Department of Computer Science and Engineering,

More information

An Image Curvature Microscope

An Image Curvature Microscope An Jean-Michel MOREL Joint work with Adina CIOMAGA and Pascal MONASSE Centre de Mathématiques et de Leurs Applications, Ecole Normale Supérieure de Cachan Séminaire Jean Serra - 70 ans April 2, 2010 Jean-Michel

More information

Expanding gait identification methods from straight to curved trajectories

Expanding gait identification methods from straight to curved trajectories Expanding gait identification methods from straight to curved trajectories Yumi Iwashita, Ryo Kurazume Kyushu University 744 Motooka Nishi-ku Fukuoka, Japan yumi@ieee.org Abstract Conventional methods

More information

This is an author-deposited version published in: Eprints ID: 4362

This is an author-deposited version published in:   Eprints ID: 4362 This is an author-deposited version published in: http://oatao.univ-toulouse.fr/ Eprints ID: 4362 To cite this document: CHIKHAOUI Oussama, GRESSIER Jérémie, GRONDIN Gilles. Assessment of the Spectral

More information

Bumptrees for Efficient Function, Constraint, and Classification Learning

Bumptrees for Efficient Function, Constraint, and Classification Learning umptrees for Efficient Function, Constraint, and Classification Learning Stephen M. Omohundro International Computer Science Institute 1947 Center Street, Suite 600 erkeley, California 94704 Abstract A

More information

Triangular Mesh Segmentation Based On Surface Normal

Triangular Mesh Segmentation Based On Surface Normal ACCV2002: The 5th Asian Conference on Computer Vision, 23--25 January 2002, Melbourne, Australia. Triangular Mesh Segmentation Based On Surface Normal Dong Hwan Kim School of Electrical Eng. Seoul Nat

More information

Model-based segmentation and recognition from range data

Model-based segmentation and recognition from range data Model-based segmentation and recognition from range data Jan Boehm Institute for Photogrammetry Universität Stuttgart Germany Keywords: range image, segmentation, object recognition, CAD ABSTRACT This

More information

Processing 3D Surface Data

Processing 3D Surface Data Processing 3D Surface Data Computer Animation and Visualisation Lecture 12 Institute for Perception, Action & Behaviour School of Informatics 3D Surfaces 1 3D surface data... where from? Iso-surfacing

More information

Snakes reparameterization for noisy images segmentation and targets tracking

Snakes reparameterization for noisy images segmentation and targets tracking Snakes reparameterization for noisy images segmentation and targets tracking Idrissi Sidi Yassine, Samir Belfkih. Lycée Tawfik Elhakim Zawiya de Noaceur, route de Marrakech, Casablanca, maroc. Laboratoire

More information

Outline. Level Set Methods. For Inverse Obstacle Problems 4. Introduction. Introduction. Martin Burger

Outline. Level Set Methods. For Inverse Obstacle Problems 4. Introduction. Introduction. Martin Burger For Inverse Obstacle Problems Martin Burger Outline Introduction Optimal Geometries Inverse Obstacle Problems & Shape Optimization Sensitivity Analysis based on Gradient Flows Numerical Methods University

More information

ing and enhancing operations for shapes represented by level sets (isosurfaces) and applied unsharp masking to the level set normals. Their approach t

ing and enhancing operations for shapes represented by level sets (isosurfaces) and applied unsharp masking to the level set normals. Their approach t Shape Deblurring with Unsharp Masking Applied to Mesh Normals Hirokazu Yagou Λ Alexander Belyaev y Daming Wei z Λ y z ; ; Shape Modeling Laboratory, University of Aizu, Aizu-Wakamatsu 965-8580 Japan fm50534,

More information

A New Algorithm for Detecting Text Line in Handwritten Documents

A New Algorithm for Detecting Text Line in Handwritten Documents A New Algorithm for Detecting Text Line in Handwritten Documents Yi Li 1, Yefeng Zheng 2, David Doermann 1, and Stefan Jaeger 1 1 Laboratory for Language and Media Processing Institute for Advanced Computer

More information

Surface Reconstruction. Gianpaolo Palma

Surface Reconstruction. Gianpaolo Palma Surface Reconstruction Gianpaolo Palma Surface reconstruction Input Point cloud With or without normals Examples: multi-view stereo, union of range scan vertices Range scans Each scan is a triangular mesh

More information

Image Enhancement Techniques for Fingerprint Identification

Image Enhancement Techniques for Fingerprint Identification March 2013 1 Image Enhancement Techniques for Fingerprint Identification Pankaj Deshmukh, Siraj Pathan, Riyaz Pathan Abstract The aim of this paper is to propose a new method in fingerprint enhancement

More information

Binary Morphological Model in Refining Local Fitting Active Contour in Segmenting Weak/Missing Edges

Binary Morphological Model in Refining Local Fitting Active Contour in Segmenting Weak/Missing Edges 0 International Conerence on Advanced Computer Science Applications and Technologies Binary Morphological Model in Reining Local Fitting Active Contour in Segmenting Weak/Missing Edges Norshaliza Kamaruddin,

More information

Implicit Active Contours Driven by Local Binary Fitting Energy

Implicit Active Contours Driven by Local Binary Fitting Energy Implicit Active Contours Driven by Local Binary Fitting Energy Chunming Li 1, Chiu-Yen Kao 2, John C. Gore 1, and Zhaohua Ding 1 1 Institute of Imaging Science 2 Department of Mathematics Vanderbilt University

More information

Grasp Recognition using a 3D Articulated Model and Infrared Images

Grasp Recognition using a 3D Articulated Model and Infrared Images Grasp Recognition using a 3D Articulated Model and Infrared Images Koichi Ogawara Institute of Industrial Science, Univ. of Tokyo, Tokyo, Japan Jun Takamatsu Institute of Industrial Science, Univ. of Tokyo,

More information

COMPUTER AND ROBOT VISION

COMPUTER AND ROBOT VISION VOLUME COMPUTER AND ROBOT VISION Robert M. Haralick University of Washington Linda G. Shapiro University of Washington A^ ADDISON-WESLEY PUBLISHING COMPANY Reading, Massachusetts Menlo Park, California

More information

BACK AND FORTH ERROR COMPENSATION AND CORRECTION METHODS FOR REMOVING ERRORS INDUCED BY UNEVEN GRADIENTS OF THE LEVEL SET FUNCTION

BACK AND FORTH ERROR COMPENSATION AND CORRECTION METHODS FOR REMOVING ERRORS INDUCED BY UNEVEN GRADIENTS OF THE LEVEL SET FUNCTION BACK AND FORTH ERROR COMPENSATION AND CORRECTION METHODS FOR REMOVING ERRORS INDUCED BY UNEVEN GRADIENTS OF THE LEVEL SET FUNCTION TODD F. DUPONT AND YINGJIE LIU Abstract. We propose a method that significantly

More information

Image Segmentation II Advanced Approaches

Image Segmentation II Advanced Approaches Image Segmentation II Advanced Approaches Jorge Jara W. 1,2 1 Department of Computer Science DCC, U. of Chile 2 SCIAN-Lab, BNI Outline 1. Segmentation I Digital image processing Segmentation basics 2.

More information

Law, AKW; Zhu, H; Lam, FK; Chan, FHY; Chan, BCB; Lu, PP

Law, AKW; Zhu, H; Lam, FK; Chan, FHY; Chan, BCB; Lu, PP Title Tumor boundary extraction in multislice MR brain images using region and contour deformation Author(s) Law, AKW; Zhu, H; Lam, FK; Chan, FHY; Chan, BCB; Lu, PP Citation International Workshop on Medical

More information

Robust and efficient 2D Skeleton Shape Representation. using Shock Graphs

Robust and efficient 2D Skeleton Shape Representation. using Shock Graphs 1/19 Robust and efficient 2D Skeleton Shape Representation using Shock Graphs Chung, In Young Lee, Kang Eui 1. Introduction An important approach to representing the structural shape of a plane region

More information

SURFACE CONSTRUCTION USING TRICOLOR MARCHING CUBES

SURFACE CONSTRUCTION USING TRICOLOR MARCHING CUBES SURFACE CONSTRUCTION USING TRICOLOR MARCHING CUBES Shaojun Liu, Jia Li Oakland University Rochester, MI 4839, USA Email: sliu2@oakland.edu, li4@oakland.edu Xiaojun Jing Beijing University of Posts and

More information

An explicit feature control approach in structural topology optimization

An explicit feature control approach in structural topology optimization th World Congress on Structural and Multidisciplinary Optimisation 07 th -2 th, June 205, Sydney Australia An explicit feature control approach in structural topology optimization Weisheng Zhang, Xu Guo

More information

10 Conclusions and future research

10 Conclusions and future research which is caused by a small narrow ridge section (about 8mm thick, much smaller than 2 times R t, where R t = 10mm ). We believe that this can be solved by examing and identifying local surface changes

More information

arxiv: v1 [math.na] 20 Sep 2016

arxiv: v1 [math.na] 20 Sep 2016 arxiv:1609.06236v1 [math.na] 20 Sep 2016 A Local Mesh Modification Strategy for Interface Problems with Application to Shape and Topology Optimization P. Gangl 1,2 and U. Langer 3 1 Doctoral Program Comp.

More information

Chapter 3. Image Processing Methods. (c) 2008 Prof. Dr. Michael M. Richter, Universität Kaiserslautern

Chapter 3. Image Processing Methods. (c) 2008 Prof. Dr. Michael M. Richter, Universität Kaiserslautern Chapter 3 Image Processing Methods The Role of Image Processing Methods (1) An image is an nxn matrix of gray or color values An image processing method is algorithm transforming such matrices or assigning

More information

coding of various parts showing different features, the possibility of rotation or of hiding covering parts of the object's surface to gain an insight

coding of various parts showing different features, the possibility of rotation or of hiding covering parts of the object's surface to gain an insight Three-Dimensional Object Reconstruction from Layered Spatial Data Michael Dangl and Robert Sablatnig Vienna University of Technology, Institute of Computer Aided Automation, Pattern Recognition and Image

More information

Multi-dimensional Image Analysis

Multi-dimensional Image Analysis Multi-dimensional Image Analysis Lucas J. van Vliet www.ph.tn.tudelft.nl/~lucas Image Analysis Paradigm scene Image formation sensor pre-processing Image enhancement Image restoration Texture filtering

More information

A Toolbox of Level Set Methods

A Toolbox of Level Set Methods A Toolbox of Level Set Methods Ian Mitchell Department of Computer Science University of British Columbia http://www.cs.ubc.ca/~mitchell mitchell@cs.ubc.ca research supported by the Natural Science and

More information

A Survey of Light Source Detection Methods

A Survey of Light Source Detection Methods A Survey of Light Source Detection Methods Nathan Funk University of Alberta Mini-Project for CMPUT 603 November 30, 2003 Abstract This paper provides an overview of the most prominent techniques for light

More information

2 Description of Complex Objects from Multiple Range Images Using an Inflating Balloon Model

2 Description of Complex Objects from Multiple Range Images Using an Inflating Balloon Model 2 Description of Complex Objects from Multiple Range Images Using an Inflating Balloon Model Yang Chen and Gérard Medioni We address the problem of constructing a complete surface model of an object using

More information

doi: /

doi: / Yiting Xie ; Anthony P. Reeves; Single 3D cell segmentation from optical CT microscope images. Proc. SPIE 934, Medical Imaging 214: Image Processing, 9343B (March 21, 214); doi:1.1117/12.243852. (214)

More information

Dynamic Obstacle Detection Based on Background Compensation in Robot s Movement Space

Dynamic Obstacle Detection Based on Background Compensation in Robot s Movement Space MATEC Web of Conferences 95 83 (7) DOI:.5/ matecconf/79583 ICMME 6 Dynamic Obstacle Detection Based on Background Compensation in Robot s Movement Space Tao Ni Qidong Li Le Sun and Lingtao Huang School

More information

Introduction. Computer Vision & Digital Image Processing. Preview. Basic Concepts from Set Theory

Introduction. Computer Vision & Digital Image Processing. Preview. Basic Concepts from Set Theory Introduction Computer Vision & Digital Image Processing Morphological Image Processing I Morphology a branch of biology concerned with the form and structure of plants and animals Mathematical morphology

More information

Level lines based disocclusion

Level lines based disocclusion Level lines based disocclusion Simon Masnou Jean-Michel Morel CEREMADE CMLA Université Paris-IX Dauphine Ecole Normale Supérieure de Cachan 75775 Paris Cedex 16, France 94235 Cachan Cedex, France Abstract

More information

Lecture 25: Bezier Subdivision. And he took unto him all these, and divided them in the midst, and laid each piece one against another: Genesis 15:10

Lecture 25: Bezier Subdivision. And he took unto him all these, and divided them in the midst, and laid each piece one against another: Genesis 15:10 Lecture 25: Bezier Subdivision And he took unto him all these, and divided them in the midst, and laid each piece one against another: Genesis 15:10 1. Divide and Conquer If we are going to build useful

More information

GRID WARPING IN TOTAL VARIATION IMAGE ENHANCEMENT METHODS. Andrey Nasonov, and Andrey Krylov

GRID WARPING IN TOTAL VARIATION IMAGE ENHANCEMENT METHODS. Andrey Nasonov, and Andrey Krylov GRID WARPING IN TOTAL VARIATION IMAGE ENHANCEMENT METHODS Andrey Nasonov, and Andrey Krylov Lomonosov Moscow State University, Moscow, Department of Computational Mathematics and Cybernetics, e-mail: nasonov@cs.msu.ru,

More information