A Discrete Search Method for Multi-modal Non-Rigid Image Registration

Size: px
Start display at page:

Download "A Discrete Search Method for Multi-modal Non-Rigid Image Registration"

Transcription

1 A Discrete Search Metho for Multi-moal on-rigi Image Registration Alexaner Shekhovtsov Juan D. García-Arteaga Tomáš Werner Center for Machine Perception, Dept. of Cybernetics Faculty of Electrical Engineering, Czech Technical University, Prague, Czech Republic Abstract We consier the problem of image matching uner the unknown statistical epenence of the signals, i.e. a signal in one image may correspon to one or more signals in the other image with ifferent probabilities. This problem is wiely known as multimoal image registration an is commonly solve by the maximization of the empirical mutual information between the images. The eformation is typically represente in a parametric form an optimization w.r.t. it is performe using graient-base methos. In contrast, we represent the eformation as a fiel of iscretize isplacements an optimize w.r.t. it using pairwise Gibbs energy minimization technique. This has potential avantage of fining goo solutions even for problems having many local minima. In experiments we emonstrate that the propose metho working on a single scale achieves comparable performance to a state-of-the-art multi-scale metho.. Introuction Image registration is efine as the process of fining the corresponence between the points in two or more images of the same scene. Its application to meical images has been well stuie 9, 3, 2, an extensively use in the fusion of complimentary ata acquire from multiple scanning evices, the comparison of images from ifferent iniviuals, an the follow-up of the changes within the same iniviual at ifferent times. When imaging biological structures, rigi eformations cannot, in general, compensate for morphological variations between iniviuals, the graual changes over time or the natural elasticity of image tissues. In these cases, espite its inherent ifficulty, non-rigi matching is the preferre image registration option. The task of estimating the non-rigi eformation that best aligns two images is usually formulate as an optimization problem that minimizes an image issimilarity criterion between the eforme (registere) images. The eformation, usually taken from a parametric family (e.g. B-Splines), is expecte to realistically moel the physical variations without introucing artifacts. As the eformation is typically escribe by a high number of parameters, the optimal solution is normally foun by a ifferential or local search metho 2. Because the objective function comparing the images uner the eformation has many local minima, it can only be rather poorly optimize by local methos. The problem becomes even more significant when aitional parameters nee to be estimate; specifically, we consier the case where the statistical epenence between intensities is unknown. Maximizing the likelihoo of the observe ata w.r.t. both the eformation an the parameters of color epenence moel naturally yiels an alternating optimization algorithm. Taking the non-parametric form of the unknown statistical color epenence it is known that this maximum likelihoo approach is equivalent to registration by maximization of empirical mutual information 4. In this article we propose to search for the best eformation aligning multi-moal meical images using, instea of local search methos, a global approximate metho. Our approach is to consier the eformation as a combination of the local translations of small blocks with the constraint that eviation of translations of neighboring blocks is restricte. Relate Work. Kim et al. 7 applie mutual information criterion to solve the stereo corresponence problem. They searche for the best next estimate of the isparity fiel (eformation) by solving a iscrete search problem in the form of pairwise Gibbs energy minimization. They showe that taking the first orer Taylor approximation of the variant of empirical mutual information with Parzen winow ensity estimator yiels the same objective as alternating the estimation of the eformation an color statistical moels in the maximum likelihoo approach. Glocker et al. 4 an Shekhovtsov et al. 8 applie linear programming relaxation to cope with ifficult iscrete search problem of the optimal 2D/3D eformation. While iscrete optimization methos were propose earlier for recovering nonrigi eformation 5,,, methos 4, 8

2 propose several approximation tricks which mae it realistically applicable for large eformations. While 4 consiere normalize mutual information as a possible measure of image issimilarity they iscuss it just very briefly. In particular, it is not clear if it was applie to local winows inepenently, like normalize cross-correlation, or otherwise. Both 4 an 8 apply similar piecewise translational approximations of the eformation (moving blocks): 4 applies soft regularization on the eformation fiel an proposes a graual multi-scale approach while 8 consiers har constraints an focuses on the efficient implementation of single-scale optimization. Hufnagel et al. 5 applie a block matching scheme with inepenent blocks, with spline post-interpolation. This approach is similar to that of 4, 8, but the blocks are matche inepenently, meaning that they must be rather large to make a goo match an this strongly restricts the class of eformations for which the approach is applicable. Roche et al. 4 presente a systematic stuy of the image registration problem from the point of view of the maximum likelihoo approach, in particular they showe that the non-parametric estimate of statistical color epenence (joint histogram) in maximum likelihoo formulation leas to a maximization of the simplest variant of the empirical mutual information. 2. Probabilistic Moel We follow the stanar computer vision approach an formulate the registration problem as a maximum a posteriori inference in a constructe probabilistic moel. A etaile treatment of image registration within maximum likelihoo approach may be foun in 4, whereas we focus on a special case. I Figure. A non-symmetric formulation of registration of image I an image J. Let I an J be two images to be registere. Let I : T C an J : T 2 C 2, where T an T 2 are sets of locations an C an C 2 are sets of signals. The eformation is represente by a mapping : T T 2. We assume that is an injective mapping, meaning that not all pixels from T 2 will have a preimage. This is illustrate in Fig.. This nonsymmetric formulation is convenient since it allows image I to be matche to a subregion of J. J Assuming that I an J are geometrically istorte observations of an unknown but equal unerlying physical escription, one erives (as in e.g. 4) that there is a statistical epenence of how a particular hien element is image in I an how it is image in J. In other wors, there is a statistical epenence between signals in the two images, which is moele as p(c,c 2 ; θ), c C, c 2 C 2, where θ is a parameter efining this istribution. Treating images I, J an the mapping as ranom variables, we moel their statistical epenence as p(i, J) =p(i, J)p(). () Assuming that pixel signals in I are conitionally inepenent given an J, we write p(i, J) = t T p(i t, J). (2) When pixel t is mappe to pixel (t) we consier that its signal I t epens only on the corresponing signal J (t) an oes not epen on the rest of signals in J. This assumption is expresse probabilistically as p(i t, J) =p(i t J (t) ; θ). (3) Prior istribution p(), etaile later, expresses how likely is eformation. In particular, it is often assume that highly curve eformations are unlikely. The maximum likelihoo estimate of eformation an parameters θ is formulate now as (,θ ) = argmax p(i t J (t) ; θ)p(). (4),θ t T We consier that sets C an C 2 are finite. This coul mean that signals in I (resp. J) woul be quantize into C (resp. C 2 ) bins, if the continuous signal spaces were one imensional. Alternatively, we can eal with multi-imensional signals by representing images I (resp. J) by C (resp. C 2 ) representative examples (much like a color palette). Our moel of the epenence p(c,c 2 ; θ) is then a Gaussian mixture moel with C C 2 Gaussians place in each possible location (c,c 2) having iagonal covariance matrix. Formally, p(c,c 2 ; θ) = θ c,c g 2 (c,c )g 2 (c 2,c 2), (5) c,c 2 where g k (c, c )= exp ρ k(c, c ) 2, Z k 2 k =, 2; (6) an ρ k (c, c ) is a metric in the space C k, k =, 2. Metrics ρ (resp. ρ 2 ) captures the important information of

3 how close are two ifferent elements of C (resp. C 2 ). A Gaussian istribution with a ifferent variance can be obtaine by changing the metric an the normalization constant accoringly. The metrics are assume to be preefine an fixe. For (5) to efine a istribution the constraint c θ,c 2 c,c =must hol. Fining the maximum 2 likelihoo estimate of parameters θ an evaluating the istribution p(c,c 2 ; θ) may be seen as very similar to Parzen winow kernel ensity estimation with Gaussian kernel. 2.. Relation to Mutual Information Mutual information was simultaneously propose as an image similarity criterion by Viola an Wells 9 an Collignon et al. 3. Since then, it has rapily become the preferre choice when registering multi-moal meical images (see the surveys by Pluim et al. 3 an Maes et al. ). Before we make any further assumptions an iscuss our approach to the optimization problem (4) we will escribe its relation to the registration by maximization of mutual information. The erivation is similar to the one presente in 4 an in 7. Taking the logarithm of (4) we can write that: =argmax log p()+max log p(i t J (t) ; θ). θ t T (7) Introucing n c,c 2 = t T δ {It=c }δ {J(t) =c 2} the maximum over θ with aitional factor may be expresse as max θ n c,c2 log p(c c 2 ; θ), (8) c C where = T = n c,c 2. Denoting ˆθ a maximizer c,c 2 of (8), the objective of (8) expresses as n c,c2 log p(c,c 2 ; ˆθ) c C p(c,c 2 ; n c,c2 log ˆθ) c C p(c 2 ; ˆθ) = n c2 log p(c 2 ; ˆθ) = Ĥ(n C,C 2 )+Ĥ(n C 2 ), (9) where n c2 = c n c,c 2 an n C,C 2 stans for the collection of numbers {n c,c 2 c C,c 2 C 2 } an n C2 for the collection of numbers { c C n c,c 2 c 2 C 2 }. The term Ĥ(n C,C 2 ) then enotes the empirical estimate of the entropy from samples represente by counts n C,C 2. The estimate is obtaine by fitting parameters of the istribution an using the sample mean instea of the expectation. It shoul be note that this is only one particular choice of how to estimate the entropy from samples an that other choices are possible. ow problem (7) can be expresse as: =argmax =argmax log p() Ĥ(n C,C 2 ()) + Ĥ(n C 2 ()) log p()+ĥ(n C ) Ĥ(n C,C 2 ()) +Ĥ(n C 2 ()) =argmax log p()+î(n C,C 2 ()), (0) where Ĥ(n C ) is the empirical entropy of n C which oes not epen on, an Î(n C,C 2 ()) is the empirical mutual information. Allowing p() to be uniform over a subset of eformations D (an zero outsie this subset), it is =argmaxî(n C,C 2 ()), () D Thus our objective may be state as the maximization of mutual information. We prefer to work with maximum likelihoo formulation because it explicitly clarifies the assumptions about the moel an allows a wier range of choices, e.g. to specify a nontrivial prior on Optimization We follow the stanar approach to optimize (4): the objective is alternatively optimize w.r.t. an θ. In this section we iscuss the etails of these two steps. Uner fixe, maximization of the likelihoo over θ is obtaine by solving (8). For this problem we apply a simple iterative algorithm 6, which in our notation reas: θc new,c 2 = θc ol n c,c 2,c 2 p(c c,c,c 2 ; θol ) g (c,c )g 2 (c 2,c 2). 2 (2) The estimate is initialize to θ c,c 2 = nc,c 2, which is easily seen to be the Parzen winow estimate, an then upate by several iterations of (2). It shoul be notice that the improvement over the initial estimate is not very significant; (2) is use instea of the simple Parzen winow estimate to be consistent with maximum likelihoo formulation. Uner fixe θ, maximizing the log-likelihoo over reas argmax log p()+ log p(i t J (t) ; θ). (3) t T We apply the metho 8 to search for the best eformation. This metho assumes that the eformation is a mapping B K, where B is the set of small blocks into which

4 the image I is subivie an K is the set of possible isplacements each block is allowe to move to (see Fig. 2). The prior on is moele as a prouct of pairwise potentials over the set E of pairs of neighboring blocks: p() = φ bb ( b, b ), (4) Z bb E where φ bb ( b, b ) is or 0 epening on whether the pair of isplacements ( b, b ) is allowe ( D) or not ( / D). This is a so-calle har moel, in which normalization constant Z oes not influence the results since all feasible eformations from D will have the same constant probability Z, an thus the optimal will not epen on Z. When moel (4) is substitute into (3), the problem can be expresse as =argmin q b ( b ) b B bb E log φ bb ( b, b ), (5) where we enote q b ( b )= t b log p(i t J b (t); θ) an switch to minimization, following the convention of energy interpretation. Problem (5) is known as a minimization of pairwise Gibbs energy which is known to be har. A linear programming relaxation approach is applie 8. For the scope of this work it is important to notice that 8 takes explicitly into account that the problem has many local minima an a certain convex relaxation of it is solve (in a certain sense suboptimally, for the sake of spee). evertheless, we refer to this metho as a global one. ote that unlike the approach of e.g. 5, there are pairwise constraints on the neighboring locations of blocks, implying that fining the best translations can only be one jointly an oes not ecompose into fining the best position for each block inepenently. After the best piece-wise translational eformation is foun, we interpolate it to obtain a smooth eformation. I b b b b Figure 2. An approximation of eformation as a piece-wise translational eformation. Image I is regularry subivie into small blocks, each block b B may be move solily by isplacement b, a pair of neighboring blocks is constraine to have small relative isplacement, thus preventing fol-overs an iscontinuities in the eformation fiel. 3. Experiments For a quantitative test we use two synthetic images shown in Fig. 3(b), which correspon to two axial cuts taken J Figure 3. A synthetically generate 2 pair of multi-moal MRI images Figure 4. Convergence plot of the log likelihoo objective function (empirical mutual information). Circles inicate optimization steps w.r.t. parameters θ of the statistical epenence of signals an boxes inicate optimization steps w.r.t. eformation. from a simulate MRI brain scan 2 with ifferent relaxation times. Figure 5(last) shows that there is a non-trivial statistical epenence between colors of the two input images when they are perfectly aligne. A typical run with secon input image eforme like it is exemplifie in Fig. 6 an both images corrupte by (0, 0. 2 ) noise (the signal space is 0, ) is as follows. The progress in log-likelihoo objective uring alternating optimization steps is shown in Fig. 4. ote that the plot starts by estimating the signal epenence parameters θ assuming that the initial eformation is the ientity. It is seen that the metho converges in a few alternating steps. Figure 5 shows the evolution of the conitional ensity estimate p(c c c ; θ). We o not show a visual comparison of registere images, since the registration is quite accurate an it is very ifficult to see the ifferences irectly. What we will be intereste in instea is to measure quantitatively how the recovere eformation iffers from the true eformation r. We teste two types of noise: Gaussian i.i.. ae to each pixel an a clutter-like noise, B(σ 2 ), which makes bright spots appear in ranom positions. Here, σ is the variance (in pixels) of the spatial Gaussian mask efining the size of the spots. We have compare our metho against an elastic metho base on free-form eformations using B-splines an local

5 Ourmetho ITK P{err a} a,px Figure 6. Left: an example of ranomly eforme an corrupte by Gaussian noise of variance The other input image is not eforme but perturbe with the noise as well. Right: the error statistics. For each value of the error the plot shows the estimate probability of the event that an error larger than that woul be encountere on a ranomly taken sample. Soli line shows our results an ashe like shows results of ITK 6, Figure 5. Estimates of the log conitional ensity log p(c c 2; θ) uring first 2 an the last iterations. Bottom: the estimate uner the groun truth eformation an the estimate for noiseless images uner groun truth eformation. P{err a} Ourmetho ITK graient escent optimization of mutual information implemente by the Image Registration Toolkit Software (ITK) 6, 7. For each recovere eformation the maximum error was measure as err = max t r t, (6) t A which correspons to the maximum eviation from the groun-truth eformation. The maximum is compute over the subset A T, which exclues the ark backgroun of the template original image. The eformation of the black backgroun is, inee, unrecoverable. By measuring this error over 00 recovere eformations, we estimate the probability P {err α} that on a ranom sample the error will be α or greater. This probability equals P {err <α} an P {err <α} is compute as empirical cumulative istribution function estimate. Results of our measurements are shown in Figs. 6 an 7. ote that although the errors may seem large (over 0 pixels for some examples), these are maximal errors, whereas average errors over pixels are shown in the Table. The registration by both methos align the images intensities very well an it is left to the regularization prior to solve any ambiguity. Actually, if the groun truth error was not known for every pixel, the error in the estimate eformation woul not be iscovere. Our current implementation takes aroun 30 secons for 37x28 pixel images a,px Figure 7. Similar to figure (6). Gaussian noise with variance 0. 2 an clutter-type noise ae to both input images. The eforme an corrupte by noise secon image is shown on the left. (0, 0. 2 ) B(2 2 ) + (0, 0. 2 ) our ITK our ITK AE Mean AE Meian AE St MOD Mean MOD Meian MOD St Table. Statistics of Angular Error (AE, egrees) an Magnitue of Difference (MOD, pixels) errors for ifferent amount of noise. Errors are compute over the area excluing the ark backgroun of the template original image. 4. Discussion an Conclusions We have presente a metho for image registration base on iscrete search techniques for mutual information an have given an empirical evience of its usefulness as an alternative for local optimization of the mutual information criterion in multi-moal images. Tests on simulate 2D meical images show that the metho is able to consistently recover large eformations in a robust manner an feasible time. When comparing the precision of the metho to

6 a specialize meical image registration system using local optimization (Image Registration Toolkit) the results show that the two systems have similar errors with ITK having a slightly better performance. One of the problems with the metho is that the approximation technique we use in the -optimization step is of course not guarantee to fin a global optimum. Furthermore, reestimating the eformation sometimes ecreases the objective. This happens because 8 fins new approximate optimal solution without taking into account the estimate from the previous step an thus is not guarantee to improve it. It shoul be note that cubic B-spline base methos have a strong avantage by using physically-inspire regularization on the eformation fiel. An open research irection is to a to the current har constraints on the eformation fiel a regularization term, e.g. a bening energy type of regularization, which may be efine using secon erivatives of the eformation fiel. It is clear that with pairwise interactions only, our iscrete energy function cannot moel secon erivatives. It is hence natural to consier higher orer Gibbs energy moels which coul impose this prior. 5. Acknowlegements The authors have been supporte by European Commission grant (DIPLECS), the Czech government grant MSM , an Czech Science Founation Project 02/07/37. The Image Registration Toolkit use to register the images by free-form eformations was use uner Licence from Ixico Lt. References Y. Boykov, O. Veksler, an R. Zabih. Fast approximate energy minimization via graph cuts. IEEE Transactions on Pattern Analysis an Machine Intelligence, 23(): , ov C. CA, K. Kwan, an R. Evans. Brainweb: online interface to a 3-D MRI simulate brain atabase. euroimage, 5:425, A. Collignon, F. Maes, D. Delaere, D. Vanermeulen, P. Suetens, an G. Marchal. Automate multi-moality image registration base on information theory. Information Processing in Meical Imaging, pages , B. Glocker,. Komoakis,. Paragios, G. Tziritas, an. avab. Inter an intra-moal eformable registration: Continuous eformations meet efficient optimal linear programming. In Information Processing in Meical Imaging, pages 8 4, H. Hufnagel, X. Pennec, G. Malanain, H. Hanels, an. Ayache. on-linear 2D an 3D registration using blockmatching an b-splines. In Bilverarbeitung fuer ie Meizin, pages , March M. Jones an D. Henerson. Maximum likelihoo kernel ensity estimation. Research Report 05/0, The Open University, UK, an University of ewcastle, UK, January J. Kim, V. Kolmogorov, an R. Zabih. Visual corresponence using energy minimization an mutual information. In International Conference on Computer Vision, pages 033 0, Komoakis an G. Tziritas. A new framework for approximate labeling via graph cuts. In International Conference on Computer Vision, pages , H. Lester an S. R. Arrige. A survey of hierarchical non-linear meical image registration. Pattern Recognition, 32():29 49, January H. Luan, F. Qi, Z. Xue, L. Chen, an D. Shen. Multimoality image registration by maximization of quantitativequalitative measure of mutual information. Pattern Recognition, 4(): , January 08. F. Maes, D. Vanermeulen, an P. Suetens. Meical image registration using mutual information. Proceeings of the IEEE, 9(0): , J. Maintz an M. Viergever. A survey of meical image registration. Meical Image Analysis, 2(): 36, J. P. W. Pluim, J. B. A. Maintz, an M. A. Viergever. Mutualinformation-base registration of meical images: a survey. Meical Imaging, IEEE Transactions on, 22(8): , A. Roche, G. Malanain, an. Ayache. Unifying maximum likelihoo approaches in meical image registration. International Journal of Imaging Systems an Technology, ():7, S. Roy an M.-A. Drouin. Mrf solutions for probabilistic optical flow formulations. In International Conference on Pattern Recognition, volume 3, Septembre D. Rueckert, L. Sonoa, C. Hayes, D. Hill, M. Leach, an D. Hawkes. on-rigi registration using free-form eformations: Application to breast MR images. IEEE Transactions on Meical Imaging, 8(8):72 72, J. A. Schnabel, D. Rueckert, M. Quist, J. M. Blackall, A. D. Castellano-Smith, T. Hartkens, G. P. Penney, W. A. Hall, H. Liu, C. L. Truwit, F. A. Gerritsen, D. L. G. Hill, an D. J. Hawkes. A generic framework for non-rigi registration base on non-uniform multi-level free-form eformations. In International Conference on Meical Image Computing an Computer-Assiste Intervention, pages , 0. 8 A. Shekhovtsov, I. Kovtun, an V. Hlavac. Efficient MRF eformation moel for non-rigi image matching. In IEEE Transactions on International Conference on Pattern Recognition, P. Viola an W. Wells, III. Alignment by maximization of mutual information. International Journal of Computer Vision, 24(2):37 54, September 997. J. Winn an. Jojic. LOCUS: Learning object classes with unsupervise segmentation. In International Conference on Computer Vision, volume, pages , 05.

Shift-map Image Registration

Shift-map Image Registration Shift-map Image Registration Linus Svärm Petter Stranmark Centre for Mathematical Sciences, Lun University {linus,petter}@maths.lth.se Abstract Shift-map image processing is a new framework base on energy

More information

Shift-map Image Registration

Shift-map Image Registration Shift-map Image Registration Svärm, Linus; Stranmark, Petter Unpublishe: 2010-01-01 Link to publication Citation for publishe version (APA): Svärm, L., & Stranmark, P. (2010). Shift-map Image Registration.

More information

Classifying Facial Expression with Radial Basis Function Networks, using Gradient Descent and K-means

Classifying Facial Expression with Radial Basis Function Networks, using Gradient Descent and K-means Classifying Facial Expression with Raial Basis Function Networks, using Graient Descent an K-means Neil Allrin Department of Computer Science University of California, San Diego La Jolla, CA 9237 nallrin@cs.ucs.eu

More information

Bayesian localization microscopy reveals nanoscale podosome dynamics

Bayesian localization microscopy reveals nanoscale podosome dynamics Nature Methos Bayesian localization microscopy reveals nanoscale poosome ynamics Susan Cox, Ewar Rosten, James Monypenny, Tijana Jovanovic-Talisman, Dylan T Burnette, Jennifer Lippincott-Schwartz, Gareth

More information

Particle Swarm Optimization Based on Smoothing Approach for Solving a Class of Bi-Level Multiobjective Programming Problem

Particle Swarm Optimization Based on Smoothing Approach for Solving a Class of Bi-Level Multiobjective Programming Problem BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 17, No 3 Sofia 017 Print ISSN: 1311-970; Online ISSN: 1314-4081 DOI: 10.1515/cait-017-0030 Particle Swarm Optimization Base

More information

Transient analysis of wave propagation in 3D soil by using the scaled boundary finite element method

Transient analysis of wave propagation in 3D soil by using the scaled boundary finite element method Southern Cross University epublications@scu 23r Australasian Conference on the Mechanics of Structures an Materials 214 Transient analysis of wave propagation in 3D soil by using the scale bounary finite

More information

Modifying ROC Curves to Incorporate Predicted Probabilities

Modifying ROC Curves to Incorporate Predicted Probabilities Moifying ROC Curves to Incorporate Preicte Probabilities Cèsar Ferri DSIC, Universitat Politècnica e València Peter Flach Department of Computer Science, University of Bristol José Hernánez-Orallo DSIC,

More information

Image Segmentation using K-means clustering and Thresholding

Image Segmentation using K-means clustering and Thresholding Image Segmentation using Kmeans clustering an Thresholing Preeti Panwar 1, Girhar Gopal 2, Rakesh Kumar 3 1M.Tech Stuent, Department of Computer Science & Applications, Kurukshetra University, Kurukshetra,

More information

Approximation with Active B-spline Curves and Surfaces

Approximation with Active B-spline Curves and Surfaces Approximation with Active B-spline Curves an Surfaces Helmut Pottmann, Stefan Leopolseer, Michael Hofer Institute of Geometry Vienna University of Technology Wiener Hauptstr. 8 10, Vienna, Austria pottmann,leopolseer,hofer

More information

Exploring Context with Deep Structured models for Semantic Segmentation

Exploring Context with Deep Structured models for Semantic Segmentation 1 Exploring Context with Deep Structure moels for Semantic Segmentation Guosheng Lin, Chunhua Shen, Anton van en Hengel, Ian Rei between an image patch an a large backgroun image region. Explicitly moeling

More information

Exploring Context with Deep Structured models for Semantic Segmentation

Exploring Context with Deep Structured models for Semantic Segmentation APPEARING IN IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, APRIL 2017. 1 Exploring Context with Deep Structure moels for Semantic Segmentation Guosheng Lin, Chunhua Shen, Anton van en

More information

THE BAYESIAN RECEIVER OPERATING CHARACTERISTIC CURVE AN EFFECTIVE APPROACH TO EVALUATE THE IDS PERFORMANCE

THE BAYESIAN RECEIVER OPERATING CHARACTERISTIC CURVE AN EFFECTIVE APPROACH TO EVALUATE THE IDS PERFORMANCE БСУ Международна конференция - 2 THE BAYESIAN RECEIVER OPERATING CHARACTERISTIC CURVE AN EFFECTIVE APPROACH TO EVALUATE THE IDS PERFORMANCE Evgeniya Nikolova, Veselina Jecheva Burgas Free University Abstract:

More information

State Indexed Policy Search by Dynamic Programming. Abstract. 1. Introduction. 2. System parameterization. Charles DuHadway

State Indexed Policy Search by Dynamic Programming. Abstract. 1. Introduction. 2. System parameterization. Charles DuHadway State Inexe Policy Search by Dynamic Programming Charles DuHaway Yi Gu 5435537 503372 December 4, 2007 Abstract We consier the reinforcement learning problem of simultaneous trajectory-following an obstacle

More information

Fast Window Based Stereo Matching for 3D Scene Reconstruction

Fast Window Based Stereo Matching for 3D Scene Reconstruction The International Arab Journal of Information Technology, Vol. 0, No. 3, May 203 209 Fast Winow Base Stereo Matching for 3D Scene Reconstruction Mohamma Mozammel Chowhury an Mohamma AL-Amin Bhuiyan Department

More information

Online Appendix to: Generalizing Database Forensics

Online Appendix to: Generalizing Database Forensics Online Appenix to: Generalizing Database Forensics KYRIACOS E. PAVLOU an RICHARD T. SNODGRASS, University of Arizona This appenix presents a step-by-step iscussion of the forensic analysis protocol that

More information

Characterizing Decoding Robustness under Parametric Channel Uncertainty

Characterizing Decoding Robustness under Parametric Channel Uncertainty Characterizing Decoing Robustness uner Parametric Channel Uncertainty Jay D. Wierer, Wahee U. Bajwa, Nigel Boston, an Robert D. Nowak Abstract This paper characterizes the robustness of ecoing uner parametric

More information

Cluster Center Initialization Method for K-means Algorithm Over Data Sets with Two Clusters

Cluster Center Initialization Method for K-means Algorithm Over Data Sets with Two Clusters Available online at www.scienceirect.com Proceia Engineering 4 (011 ) 34 38 011 International Conference on Avances in Engineering Cluster Center Initialization Metho for K-means Algorithm Over Data Sets

More information

6 Gradient Descent. 6.1 Functions

6 Gradient Descent. 6.1 Functions 6 Graient Descent In this topic we will iscuss optimizing over general functions f. Typically the function is efine f : R! R; that is its omain is multi-imensional (in this case -imensional) an output

More information

Coupling the User Interfaces of a Multiuser Program

Coupling the User Interfaces of a Multiuser Program Coupling the User Interfaces of a Multiuser Program PRASUN DEWAN University of North Carolina at Chapel Hill RAJIV CHOUDHARY Intel Corporation We have evelope a new moel for coupling the user-interfaces

More information

Estimating Velocity Fields on a Freeway from Low Resolution Video

Estimating Velocity Fields on a Freeway from Low Resolution Video Estimating Velocity Fiels on a Freeway from Low Resolution Vieo Young Cho Department of Statistics University of California, Berkeley Berkeley, CA 94720-3860 Email: young@stat.berkeley.eu John Rice Department

More information

A Comparative Evaluation of Iris and Ocular Recognition Methods on Challenging Ocular Images

A Comparative Evaluation of Iris and Ocular Recognition Methods on Challenging Ocular Images A Comparative Evaluation of Iris an Ocular Recognition Methos on Challenging Ocular Images Vishnu Naresh Boeti Carnegie Mellon University Pittsburgh, PA 523 naresh@cmu.eu Jonathon M Smereka Carnegie Mellon

More information

Discrete Markov Image Modeling and Inference on the Quadtree

Discrete Markov Image Modeling and Inference on the Quadtree 390 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 9, NO. 3, MARCH 2000 Discrete Markov Image Moeling an Inference on the Quatree Jean-Marc Laferté, Patrick Pérez, an Fabrice Heitz Abstract Noncasual Markov

More information

Research Article Inviscid Uniform Shear Flow past a Smooth Concave Body

Research Article Inviscid Uniform Shear Flow past a Smooth Concave Body International Engineering Mathematics Volume 04, Article ID 46593, 7 pages http://x.oi.org/0.55/04/46593 Research Article Invisci Uniform Shear Flow past a Smooth Concave Boy Abullah Mura Department of

More information

Vision-based Multi-Robot Simultaneous Localization and Mapping

Vision-based Multi-Robot Simultaneous Localization and Mapping Vision-base Multi-Robot Simultaneous Localization an Mapping Hassan Hajjiab an Robert Laganière VIVA Research Lab School of Information Technology an Engineering University of Ottawa, Ottawa, Canaa, K1N

More information

arxiv: v2 [cs.lg] 22 Jan 2019

arxiv: v2 [cs.lg] 22 Jan 2019 Spatial Variational Auto-Encoing via Matrix-Variate Normal Distributions Zhengyang Wang Hao Yuan Shuiwang Ji arxiv:1705.06821v2 [cs.lg] 22 Jan 2019 Abstract The key iea of variational auto-encoers (VAEs)

More information

1 Surprises in high dimensions

1 Surprises in high dimensions 1 Surprises in high imensions Our intuition about space is base on two an three imensions an can often be misleaing in high imensions. It is instructive to analyze the shape an properties of some basic

More information

Non-homogeneous Generalization in Privacy Preserving Data Publishing

Non-homogeneous Generalization in Privacy Preserving Data Publishing Non-homogeneous Generalization in Privacy Preserving Data Publishing W. K. Wong, Nios Mamoulis an Davi W. Cheung Department of Computer Science, The University of Hong Kong Pofulam Roa, Hong Kong {wwong2,nios,cheung}@cs.hu.h

More information

A Convex Clustering-based Regularizer for Image Segmentation

A Convex Clustering-based Regularizer for Image Segmentation Vision, Moeling, an Visualization (2015) D. Bommes, T. Ritschel an T. Schultz (Es.) A Convex Clustering-base Regularizer for Image Segmentation Benjamin Hell (TU Braunschweig), Marcus Magnor (TU Braunschweig)

More information

SURVIVABLE IP OVER WDM: GUARANTEEEING MINIMUM NETWORK BANDWIDTH

SURVIVABLE IP OVER WDM: GUARANTEEEING MINIMUM NETWORK BANDWIDTH SURVIVABLE IP OVER WDM: GUARANTEEEING MINIMUM NETWORK BANDWIDTH Galen H Sasaki Dept Elec Engg, U Hawaii 2540 Dole Street Honolul HI 96822 USA Ching-Fong Su Fuitsu Laboratories of America 595 Lawrence Expressway

More information

A Duality Based Approach for Realtime TV-L 1 Optical Flow

A Duality Based Approach for Realtime TV-L 1 Optical Flow A Duality Base Approach for Realtime TV-L 1 Optical Flow C. Zach 1, T. Pock 2, an H. Bischof 2 1 VRVis Research Center 2 Institute for Computer Graphics an Vision, TU Graz Abstract. Variational methos

More information

FINDING OPTICAL DISPERSION OF A PRISM WITH APPLICATION OF MINIMUM DEVIATION ANGLE MEASUREMENT METHOD

FINDING OPTICAL DISPERSION OF A PRISM WITH APPLICATION OF MINIMUM DEVIATION ANGLE MEASUREMENT METHOD Warsaw University of Technology Faculty of Physics Physics Laboratory I P Joanna Konwerska-Hrabowska 6 FINDING OPTICAL DISPERSION OF A PRISM WITH APPLICATION OF MINIMUM DEVIATION ANGLE MEASUREMENT METHOD.

More information

Almost Disjunct Codes in Large Scale Multihop Wireless Network Media Access Control

Almost Disjunct Codes in Large Scale Multihop Wireless Network Media Access Control Almost Disjunct Coes in Large Scale Multihop Wireless Network Meia Access Control D. Charles Engelhart Anan Sivasubramaniam Penn. State University University Park PA 682 engelhar,anan @cse.psu.eu Abstract

More information

Feature Extraction and Rule Classification Algorithm of Digital Mammography based on Rough Set Theory

Feature Extraction and Rule Classification Algorithm of Digital Mammography based on Rough Set Theory Feature Extraction an Rule Classification Algorithm of Digital Mammography base on Rough Set Theory Aboul Ella Hassanien Jafar M. H. Ali. Kuwait University, Faculty of Aministrative Science, Quantitative

More information

EXACT SIMULATION OF A BOOLEAN MODEL

EXACT SIMULATION OF A BOOLEAN MODEL Original Research Paper oi:10.5566/ias.v32.p101-105 EXACT SIMULATION OF A BOOLEAN MODEL CHRISTIAN LANTUÉJOULB MinesParisTech 35 rue Saint-Honoré 77305 Fontainebleau France e-mail: christian.lantuejoul@mines-paristech.fr

More information

Multimodal Stereo Image Registration for Pedestrian Detection

Multimodal Stereo Image Registration for Pedestrian Detection Multimoal Stereo Image Registration for Peestrian Detection Stephen Krotosky an Mohan Trivei Abstract This paper presents an approach for the registration of multimoal imagery for peestrian etection when

More information

PAMPAS: Real-Valued Graphical Models for Computer Vision. Abstract. 1. Introduction. Anonymous

PAMPAS: Real-Valued Graphical Models for Computer Vision. Abstract. 1. Introduction. Anonymous PAMPAS: Real-Value Graphical Moels for Computer Vision Anonymous Abstract Probabilistic moels have been aopte for many computer vision applications, however inference in high-imensional spaces remains

More information

Open Access Adaptive Image Enhancement Algorithm with Complex Background

Open Access Adaptive Image Enhancement Algorithm with Complex Background Sen Orers for Reprints to reprints@benthamscience.ae 594 The Open Cybernetics & Systemics Journal, 205, 9, 594-600 Open Access Aaptive Image Enhancement Algorithm with Complex Bacgroun Zhang Pai * epartment

More information

I. Introuction With the evolution of imaging technology, an increasing number of image moalities becomes available. In remote sensing, sensors are use

I. Introuction With the evolution of imaging technology, an increasing number of image moalities becomes available. In remote sensing, sensors are use A multivalue image wavelet representation base on multiscale funamental forms P. Scheuners Vision Lab, Department ofphysics, University ofantwerp, Groenenborgerlaan 7, 00 Antwerpen, Belgium Tel.: +3/3/8

More information

A shortest path algorithm in multimodal networks: a case study with time varying costs

A shortest path algorithm in multimodal networks: a case study with time varying costs A shortest path algorithm in multimoal networks: a case stuy with time varying costs Daniela Ambrosino*, Anna Sciomachen* * Department of Economics an Quantitative Methos (DIEM), University of Genoa Via

More information

Figure 1: Schematic of an SEM [source: ]

Figure 1: Schematic of an SEM [source:   ] EECI Course: -9 May 1 by R. Sanfelice Hybri Control Systems Eelco van Horssen E.P.v.Horssen@tue.nl Project: Scanning Electron Microscopy Introuction In Scanning Electron Microscopy (SEM) a (bunle) beam

More information

A Versatile Model-Based Visibility Measure for Geometric Primitives

A Versatile Model-Based Visibility Measure for Geometric Primitives A Versatile Moel-Base Visibility Measure for Geometric Primitives Marc M. Ellenrieer 1,LarsKrüger 1, Dirk Stößel 2, an Marc Hanheie 2 1 DaimlerChrysler AG, Research & Technology, 89013 Ulm, Germany 2 Faculty

More information

Rough Set Approach for Classification of Breast Cancer Mammogram Images

Rough Set Approach for Classification of Breast Cancer Mammogram Images Rough Set Approach for Classification of Breast Cancer Mammogram Images Aboul Ella Hassanien Jafar M. H. Ali. Kuwait University, Faculty of Aministrative Science, Quantitative Methos an Information Systems

More information

A Neural Network Model Based on Graph Matching and Annealing :Application to Hand-Written Digits Recognition

A Neural Network Model Based on Graph Matching and Annealing :Application to Hand-Written Digits Recognition ITERATIOAL JOURAL OF MATHEMATICS AD COMPUTERS I SIMULATIO A eural etwork Moel Base on Graph Matching an Annealing :Application to Han-Written Digits Recognition Kyunghee Lee Abstract We present a neural

More information

New Geometric Interpretation and Analytic Solution for Quadrilateral Reconstruction

New Geometric Interpretation and Analytic Solution for Quadrilateral Reconstruction New Geometric Interpretation an Analytic Solution for uarilateral Reconstruction Joo-Haeng Lee Convergence Technology Research Lab ETRI Daejeon, 305 777, KOREA Abstract A new geometric framework, calle

More information

Sampling the Disparity Space Image

Sampling the Disparity Space Image Accepte for publication, June 2003. To appear in IEEE Transactions on Pattern Analysis an Machine Intelligence (PAMI) Sampling the Disparity Space Image Richar Szeliski Daniel Scharstein Microsoft Research

More information

Selection Strategies for Initial Positions and Initial Velocities in Multi-optima Particle Swarms

Selection Strategies for Initial Positions and Initial Velocities in Multi-optima Particle Swarms ACM, 2011. This is the author s version of the work. It is poste here by permission of ACM for your personal use. Not for reistribution. The efinitive version was publishe in Proceeings of the 13th Annual

More information

Lecture 1 September 4, 2013

Lecture 1 September 4, 2013 CS 84r: Incentives an Information in Networks Fall 013 Prof. Yaron Singer Lecture 1 September 4, 013 Scribe: Bo Waggoner 1 Overview In this course we will try to evelop a mathematical unerstaning for the

More information

Fast Fractal Image Compression using PSO Based Optimization Techniques

Fast Fractal Image Compression using PSO Based Optimization Techniques Fast Fractal Compression using PSO Base Optimization Techniques A.Krishnamoorthy Visiting faculty Department Of ECE University College of Engineering panruti rishpci89@gmail.com S.Buvaneswari Visiting

More information

A Plane Tracker for AEC-automation Applications

A Plane Tracker for AEC-automation Applications A Plane Tracker for AEC-automation Applications Chen Feng *, an Vineet R. Kamat Department of Civil an Environmental Engineering, University of Michigan, Ann Arbor, USA * Corresponing author (cforrest@umich.eu)

More information

Learning convex bodies is hard

Learning convex bodies is hard Learning convex boies is har Navin Goyal Microsoft Research Inia navingo@microsoftcom Luis Raemacher Georgia Tech lraemac@ccgatecheu Abstract We show that learning a convex boy in R, given ranom samples

More information

Evolutionary Optimisation Methods for Template Based Image Registration

Evolutionary Optimisation Methods for Template Based Image Registration Evolutionary Optimisation Methos for Template Base Image Registration Lukasz A Machowski, Tshilizi Marwala School of Electrical an Information Engineering University of Witwatersran, Johannesburg, South

More information

Using Vector and Raster-Based Techniques in Categorical Map Generalization

Using Vector and Raster-Based Techniques in Categorical Map Generalization Thir ICA Workshop on Progress in Automate Map Generalization, Ottawa, 12-14 August 1999 1 Using Vector an Raster-Base Techniques in Categorical Map Generalization Beat Peter an Robert Weibel Department

More information

BIJECTIONS FOR PLANAR MAPS WITH BOUNDARIES

BIJECTIONS FOR PLANAR MAPS WITH BOUNDARIES BIJECTIONS FOR PLANAR MAPS WITH BOUNDARIES OLIVIER BERNARDI AND ÉRIC FUSY Abstract. We present bijections for planar maps with bounaries. In particular, we obtain bijections for triangulations an quarangulations

More information

Study of Network Optimization Method Based on ACL

Study of Network Optimization Method Based on ACL Available online at www.scienceirect.com Proceia Engineering 5 (20) 3959 3963 Avance in Control Engineering an Information Science Stuy of Network Optimization Metho Base on ACL Liu Zhian * Department

More information

Discriminative Filters for Depth from Defocus

Discriminative Filters for Depth from Defocus Discriminative Filters for Depth from Defocus Fahim Mannan an Michael S. Langer School of Computer Science, McGill University Montreal, Quebec HA 0E9, Canaa. {fmannan, langer}@cim.mcgill.ca Abstract Depth

More information

Learning Subproblem Complexities in Distributed Branch and Bound

Learning Subproblem Complexities in Distributed Branch and Bound Learning Subproblem Complexities in Distribute Branch an Boun Lars Otten Department of Computer Science University of California, Irvine lotten@ics.uci.eu Rina Dechter Department of Computer Science University

More information

Dense Disparity Estimation in Ego-motion Reduced Search Space

Dense Disparity Estimation in Ego-motion Reduced Search Space Dense Disparity Estimation in Ego-motion Reuce Search Space Luka Fućek, Ivan Marković, Igor Cvišić, Ivan Petrović University of Zagreb, Faculty of Electrical Engineering an Computing, Croatia (e-mail:

More information

A Framework for Dialogue Detection in Movies

A Framework for Dialogue Detection in Movies A Framework for Dialogue Detection in Movies Margarita Kotti, Constantine Kotropoulos, Bartosz Ziólko, Ioannis Pitas, an Vassiliki Moschou Department of Informatics, Aristotle University of Thessaloniki

More information

NET Institute*

NET Institute* NET Institute* www.netinst.org Working Paper #08-24 October 2008 Computer Virus Propagation in a Network Organization: The Interplay between Social an Technological Networks Hsing Kenny Cheng an Hong Guo

More information

Real Time On Board Stereo Camera Pose through Image Registration*

Real Time On Board Stereo Camera Pose through Image Registration* 28 IEEE Intelligent Vehicles Symposium Einhoven University of Technology Einhoven, The Netherlans, June 4-6, 28 Real Time On Boar Stereo Camera Pose through Image Registration* Fai Dornaika French National

More information

P. Fua and Y. G. Leclerc. SRI International. 333 Ravenswood Avenue, Menlo Park, CA

P. Fua and Y. G. Leclerc. SRI International. 333 Ravenswood Avenue, Menlo Park, CA Moel Driven Ege Detection P. Fua an Y. G. Leclerc SI International 333 avenswoo Avenue, Menlo Park, CA 9425 (fua@ai.sri.com leclerc@ai.sri.com) Machine Vision an Applications, 3, 199 Abstract Stanar ege

More information

Finite Automata Implementations Considering CPU Cache J. Holub

Finite Automata Implementations Considering CPU Cache J. Holub Finite Automata Implementations Consiering CPU Cache J. Holub The finite automata are mathematical moels for finite state systems. More general finite automaton is the noneterministic finite automaton

More information

MORA: a Movement-Based Routing Algorithm for Vehicle Ad Hoc Networks

MORA: a Movement-Based Routing Algorithm for Vehicle Ad Hoc Networks : a Movement-Base Routing Algorithm for Vehicle A Hoc Networks Fabrizio Granelli, Senior Member, Giulia Boato, Member, an Dzmitry Kliazovich, Stuent Member Abstract Recent interest in car-to-car communications

More information

Skyline Community Search in Multi-valued Networks

Skyline Community Search in Multi-valued Networks Syline Community Search in Multi-value Networs Rong-Hua Li Beijing Institute of Technology Beijing, China lironghuascut@gmail.com Jeffrey Xu Yu Chinese University of Hong Kong Hong Kong, China yu@se.cuh.eu.h

More information

Regularized Laplacian Zero Crossings as Optimal Edge Integrators

Regularized Laplacian Zero Crossings as Optimal Edge Integrators Regularize aplacian Zero rossings as Optimal Ege Integrators R. KIMME A.M. BRUKSTEIN Department of omputer Science Technion Israel Institute of Technology Technion ity, Haifa 32, Israel Abstract We view

More information

Multi-camera tracking algorithm study based on information fusion

Multi-camera tracking algorithm study based on information fusion International Conference on Avance Electronic Science an Technolog (AEST 016) Multi-camera tracking algorithm stu base on information fusion a Guoqiang Wang, Shangfu Li an Xue Wen School of Electronic

More information

A fast embedded selection approach for color texture classification using degraded LBP

A fast embedded selection approach for color texture classification using degraded LBP A fast embee selection approach for color texture classification using egrae A. Porebski, N. Vanenbroucke an D. Hama Laboratoire LISIC - EA 4491 - Université u Littoral Côte Opale - 50, rue Ferinan Buisson

More information

Using the disparity space to compute occupancy grids from stereo-vision

Using the disparity space to compute occupancy grids from stereo-vision The 2010 IEEE/RSJ International Conference on Intelligent Robots an Systems October 18-22, 2010, Taipei, Taiwan Using the isparity space to compute occupancy gris from stereo-vision Mathias Perrollaz,

More information

The Panum Proxy algorithm for dense stereo matching over a volume of interest

The Panum Proxy algorithm for dense stereo matching over a volume of interest The Panum Proxy algorithm for ense stereo matching over a volume of interest A. Agarwal an A. Blake Microsoft Research Lt. 7 J J Thomson Ave, Cambrige, CB3 FB, UK http://research.microsoft.com/vision/cambrige

More information

EFFICIENT STEREO MATCHING BASED ON A NEW CONFIDENCE METRIC. Won-Hee Lee, Yumi Kim, and Jong Beom Ra

EFFICIENT STEREO MATCHING BASED ON A NEW CONFIDENCE METRIC. Won-Hee Lee, Yumi Kim, and Jong Beom Ra th European Signal Processing Conference (EUSIPCO ) Bucharest, omania, August 7-3, EFFICIENT STEEO MATCHING BASED ON A NEW CONFIDENCE METIC Won-Hee Lee, Yumi Kim, an Jong Beom a Department of Electrical

More information

Robust Camera Calibration for an Autonomous Underwater Vehicle

Robust Camera Calibration for an Autonomous Underwater Vehicle obust Camera Calibration for an Autonomous Unerwater Vehicle Matthew Bryant, Davi Wettergreen *, Samer Aballah, Alexaner Zelinsky obotic Systems Laboratory Department of Engineering, FEIT Department of

More information

Advanced method of NC programming for 5-axis machining

Advanced method of NC programming for 5-axis machining Available online at www.scienceirect.com Proceia CIRP (0 ) 0 07 5 th CIRP Conference on High Performance Cutting 0 Avance metho of NC programming for 5-axis machining Sergej N. Grigoriev a, A.A. Kutin

More information

A PARALLEL APPROACH FOR ALIGNMENT OF MULTI- MODAL GRID-BASED DATA

A PARALLEL APPROACH FOR ALIGNMENT OF MULTI- MODAL GRID-BASED DATA ISBN: 978-97-894-97-3 009 IADIS A PARALLEL APPROACH FOR ALIGNMENT OF MULTI- MODAL GRID-BASED DATA Egor Dranischnikow, Elmar Schömer Institute of Computer Science, Johannes Gutenberg-University, Mainz Ulrich

More information

Interference and diffraction are the important phenomena that distinguish. Interference and Diffraction

Interference and diffraction are the important phenomena that distinguish. Interference and Diffraction C H A P T E R 33 Interference an Diffraction 33- Phase Difference an Coherence 33-2 Interference in Thin Films 33-3 Two-Slit Interference Pattern 33-4 Diffraction Pattern of a Single Slit * 33-5 Using

More information

Multilevel Linear Dimensionality Reduction using Hypergraphs for Data Analysis

Multilevel Linear Dimensionality Reduction using Hypergraphs for Data Analysis Multilevel Linear Dimensionality Reuction using Hypergraphs for Data Analysis Haw-ren Fang Department of Computer Science an Engineering University of Minnesota; Minneapolis, MN 55455 hrfang@csumneu ABSTRACT

More information

Comparative Study of Projection/Back-projection Schemes in Cryo-EM Tomography

Comparative Study of Projection/Back-projection Schemes in Cryo-EM Tomography Comparative Stuy of Projection/Back-projection Schemes in Cryo-EM Tomography Yu Liu an Jong Chul Ye Department of BioSystems Korea Avance Institute of Science an Technology, Daejeon, Korea ABSTRACT In

More information

CAMERAS AND GRAVITY: ESTIMATING PLANAR OBJECT ORIENTATION. Zhaoyin Jia, Andrew Gallagher, Tsuhan Chen

CAMERAS AND GRAVITY: ESTIMATING PLANAR OBJECT ORIENTATION. Zhaoyin Jia, Andrew Gallagher, Tsuhan Chen CAMERAS AND GRAVITY: ESTIMATING PLANAR OBJECT ORIENTATION Zhaoyin Jia, Anrew Gallagher, Tsuhan Chen School of Electrical an Computer Engineering, Cornell University ABSTRACT Photography on a mobile camera

More information

EXTRACTION OF MAIN AND SECONDARY ROADS IN VHR IMAGES USING A HIGHER-ORDER PHASE FIELD MODEL

EXTRACTION OF MAIN AND SECONDARY ROADS IN VHR IMAGES USING A HIGHER-ORDER PHASE FIELD MODEL EXTRACTION OF MAIN AND SECONDARY ROADS IN VHR IMAGES USING A HIGHER-ORDER PHASE FIELD MODEL Ting Peng a,b, Ian H. Jermyn b, Véronique Prinet a, Josiane Zerubia b a LIAMA & PR, Institute of Automation,

More information

A Revised Simplex Search Procedure for Stochastic Simulation Response Surface Optimization

A Revised Simplex Search Procedure for Stochastic Simulation Response Surface Optimization 272 INFORMS Journal on Computing 0899-1499 100 1204-0272 $05.00 Vol. 12, No. 4, Fall 2000 2000 INFORMS A Revise Simplex Search Proceure for Stochastic Simulation Response Surface Optimization DAVID G.

More information

Animated Surface Pasting

Animated Surface Pasting Animate Surface Pasting Clara Tsang an Stephen Mann Computing Science Department University of Waterloo 200 University Ave W. Waterloo, Ontario Canaa N2L 3G1 e-mail: clftsang@cgl.uwaterloo.ca, smann@cgl.uwaterloo.ca

More information

Handling missing values in kernel methods with application to microbiology data

Handling missing values in kernel methods with application to microbiology data an Machine Learning. Bruges (Belgium), 24-26 April 2013, i6oc.com publ., ISBN 978-2-87419-081-0. Available from http://www.i6oc.com/en/livre/?gcoi=28001100131010. Hanling missing values in kernel methos

More information

UNIT 9 INTERFEROMETRY

UNIT 9 INTERFEROMETRY UNIT 9 INTERFEROMETRY Structure 9.1 Introuction Objectives 9. Interference of Light 9.3 Light Sources for 9.4 Applie to Flatness Testing 9.5 in Testing of Surface Contour an Measurement of Height 9.6 Interferometers

More information

Design of Policy-Aware Differentially Private Algorithms

Design of Policy-Aware Differentially Private Algorithms Design of Policy-Aware Differentially Private Algorithms Samuel Haney Due University Durham, NC, USA shaney@cs.ue.eu Ashwin Machanavajjhala Due University Durham, NC, USA ashwin@cs.ue.eu Bolin Ding Microsoft

More information

An Investigation in the Use of Vehicle Reidentification for Deriving Travel Time and Travel Time Distributions

An Investigation in the Use of Vehicle Reidentification for Deriving Travel Time and Travel Time Distributions An Investigation in the Use of Vehicle Reientification for Deriving Travel Time an Travel Time Distributions Carlos Sun Department of Civil an Environmental Engineering, University of Missouri-Columbia,

More information

Tight Wavelet Frame Decomposition and Its Application in Image Processing

Tight Wavelet Frame Decomposition and Its Application in Image Processing ITB J. Sci. Vol. 40 A, No., 008, 151-165 151 Tight Wavelet Frame Decomposition an Its Application in Image Processing Mahmu Yunus 1, & Henra Gunawan 1 1 Analysis an Geometry Group, FMIPA ITB, Banung Department

More information

Generalized Edge Coloring for Channel Assignment in Wireless Networks

Generalized Edge Coloring for Channel Assignment in Wireless Networks TR-IIS-05-021 Generalize Ege Coloring for Channel Assignment in Wireless Networks Chun-Chen Hsu, Pangfeng Liu, Da-Wei Wang, Jan-Jan Wu December 2005 Technical Report No. TR-IIS-05-021 http://www.iis.sinica.eu.tw/lib/techreport/tr2005/tr05.html

More information

Hybrid Spline-based Multimodal Registration using a Local Measure for Mutual Information

Hybrid Spline-based Multimodal Registration using a Local Measure for Mutual Information Hybrid Spline-based Multimodal Registration using a Local Measure for Mutual Information Andreas Biesdorf 1, Stefan Wörz 1, Hans-Jürgen Kaiser 2, Karl Rohr 1 1 University of Heidelberg, BIOQUANT, IPMB,

More information

Bends, Jogs, And Wiggles for Railroad Tracks and Vehicle Guide Ways

Bends, Jogs, And Wiggles for Railroad Tracks and Vehicle Guide Ways Ben, Jogs, An Wiggles for Railroa Tracks an Vehicle Guie Ways Louis T. Klauer Jr., PhD, PE. Work Soft 833 Galer Dr. Newtown Square, PA 19073 lklauer@wsof.com Preprint, June 4, 00 Copyright 00 by Louis

More information

A multiple wavelength unwrapping algorithm for digital fringe profilometry based on spatial shift estimation

A multiple wavelength unwrapping algorithm for digital fringe profilometry based on spatial shift estimation University of Wollongong Research Online Faculty of Engineering an Information Sciences - Papers: Part A Faculty of Engineering an Information Sciences 214 A multiple wavelength unwrapping algorithm for

More information

Solution Representation for Job Shop Scheduling Problems in Ant Colony Optimisation

Solution Representation for Job Shop Scheduling Problems in Ant Colony Optimisation Solution Representation for Job Shop Scheuling Problems in Ant Colony Optimisation James Montgomery, Carole Faya 2, an Sana Petrovic 2 Faculty of Information & Communication Technologies, Swinburne University

More information

Refinement of scene depth from stereo camera ego-motion parameters

Refinement of scene depth from stereo camera ego-motion parameters Refinement of scene epth from stereo camera ego-motion parameters Piotr Skulimowski, Pawel Strumillo An algorithm for refinement of isparity (epth) map from stereoscopic sequences is propose. The metho

More information

The Reconstruction of Graphs. Dhananjay P. Mehendale Sir Parashurambhau College, Tilak Road, Pune , India. Abstract

The Reconstruction of Graphs. Dhananjay P. Mehendale Sir Parashurambhau College, Tilak Road, Pune , India. Abstract The Reconstruction of Graphs Dhananay P. Mehenale Sir Parashurambhau College, Tila Roa, Pune-4030, Inia. Abstract In this paper we iscuss reconstruction problems for graphs. We evelop some new ieas lie

More information

Exercises of PIV. incomplete draft, version 0.0. October 2009

Exercises of PIV. incomplete draft, version 0.0. October 2009 Exercises of PIV incomplete raft, version 0.0 October 2009 1 Images Images are signals efine in 2D or 3D omains. They can be vector value (e.g., color images), real (monocromatic images), complex or binary

More information

Module13:Interference-I Lecture 13: Interference-I

Module13:Interference-I Lecture 13: Interference-I Moule3:Interference-I Lecture 3: Interference-I Consier a situation where we superpose two waves. Naively, we woul expect the intensity (energy ensity or flux) of the resultant to be the sum of the iniviual

More information

Software Reliability Modeling and Cost Estimation Incorporating Testing-Effort and Efficiency

Software Reliability Modeling and Cost Estimation Incorporating Testing-Effort and Efficiency Software Reliability Moeling an Cost Estimation Incorporating esting-effort an Efficiency Chin-Yu Huang, Jung-Hua Lo, Sy-Yen Kuo, an Michael R. Lyu -+ Department of Electrical Engineering Computer Science

More information

CONTENT-BASED RETRIEVAL OF DEFECT IMAGES. Jukka Iivarinen and Jussi Pakkanen

CONTENT-BASED RETRIEVAL OF DEFECT IMAGES. Jukka Iivarinen and Jussi Pakkanen Proceeings of ACIVS 2002 (Avance Concepts for Intelligent Vision Systems), Ghent, Belgium, September 9-11, 2002 CONTENT-BASED RETRIEVAL OF DEFECT IMAGES Jukka Iivarinen an Jussi Pakkanen jukka.iivarinen@hut.fi,

More information

An Algorithm for Building an Enterprise Network Topology Using Widespread Data Sources

An Algorithm for Building an Enterprise Network Topology Using Widespread Data Sources An Algorithm for Builing an Enterprise Network Topology Using Wiesprea Data Sources Anton Anreev, Iurii Bogoiavlenskii Petrozavosk State University Petrozavosk, Russia {anreev, ybgv}@cs.petrsu.ru Abstract

More information

Interior Permanent Magnet Synchronous Motor (IPMSM) Adaptive Genetic Parameter Estimation

Interior Permanent Magnet Synchronous Motor (IPMSM) Adaptive Genetic Parameter Estimation Interior Permanent Magnet Synchronous Motor (IPMSM) Aaptive Genetic Parameter Estimation Java Rezaie, Mehi Gholami, Reza Firouzi, Tohi Alizaeh, Karim Salashoor Abstract - Interior permanent magnet synchronous

More information

A half-scan error reduction based algorithm for cone-beam CT

A half-scan error reduction based algorithm for cone-beam CT Journal of X-Ray Science an Technology 12 (2004) 73 82 73 IOS Press A half-scan error reuction base algorithm for cone-beam CT Kai Zeng a, Zhiqiang Chen a, Li Zhang a an Ge Wang a,b a Department of Engineering

More information

Texture Defect Detection System with Image Deflection Compensation

Texture Defect Detection System with Image Deflection Compensation Texture Defect Detection System with Image Deflection Compensation CHUN-CHENG LIN CHENG-YU YEH Department of Electrical Engineering National Chin-Yi University of Technology 35, Lane 15, Sec. 1, Chungshan

More information