TREE SPECIES CLASSIFICATION USING RADIOMETRY, TEXTURE AND SHAPE BASED FEATURES
|
|
- Sylvia Eaton
- 6 years ago
- Views:
Transcription
1 TREE SPECIES CLASSIFICATIO USIG RADIOMETRY, TEXTURE AD SHAPE BASED FEATURES Maria S. Kulikova 1, Meena Mani 1, Anuj Srivastava 2, Xavier Descomes 1, Josiane Zeruia 1 1 Ariana Research Group, IRIA/I3S Sophia Antipolis Ceex France firstname.lastname@inria.fr, mmani@ucla.eu to join Meena Mani 2 Department of Statistics Floria State University Tallahassee, FL anuj@ani.stat.fsu.eu ABSTRACT We consier the prolem of tree species classification from high resolution aerial images ase on raiometry, texture an a shape moeling. We use the notion of shape space propose y Klassen et al., which provies a shape escription invariant to translation, rotation an scaling. The shape features are extracte within a geoesic istance in the shape space. We then perform a classification using a SVM approach. We are ale to show that the shape escriptors improve the classification performance relative to a classifier ase on raiometric an textural escriptors alone. We otain these results using high resolution Colour InfraRe (CIR) aerial images provie y the Sweish University of Agricultural Sciences. The image viewpoint is close to the nair, i.e. the tree crowns are seen from aove. 1. ITRODUCTIO Interest in applying remote sensing to forests goes ack to the 1920s when aerial photographs were first use to assess forest inventory. Remote sensing is now wiely employe in forest management where the aerial information is comine with measurements taken on the groun to stuy the ioiversity of the forest ecosystem. The methos evelope in forest image processing aim to facilitate the task of forest inventory an assessment. The most useful parameters otaine from aerial images an groun measurements are ensity (of planting), age of trees, stem volume, tree species composition, an information such as iotops an haitats that have ecological value. In orer to otain information aout the iversity of the forest species an stem volume, the classification of tree crowns into species is necessary. Prior to classification, we nee to segment the image into iniviual tree crowns. Segmentation techniques such as template matching, ege etection an others that have een employe for this application are iscusse in [5], [13]. A few approaches have een propose to classify the trees into species. One metho is the Signature Generation Process where for every crown extracte, a class of signatures is create from the multispectral ata of initial image, (cf [10]). A likelihoo maximisation technique is use to lael the crown. Some crowns with signatures too far from a succesful match remain unclassifie. In another stuy [6], tree crowns are manually elineate to avoi ias ue to a etection. 50 trees for each class verifie against the groun truth are selecte. 7 parameters such as the multispectral average of pixels in the crown (average on each of the ans) an the illuminate part of the crown, as well as the multispectral value of the tree crown (most lighte pixel) are then compute. Spectral signatures of a crown or a region within a crown are evelope y The work of the first author has een partly supporte y a grant of the French Government. This work has een conucte within the IRIA ARC Moe e Vie joint research program ( an IRIA/FSU associate team SHAPES ( comining means an covariance patterns. The istinct characteristics allow us to regroup elineate crowns in foreste populations [7]. There are some limitations with this metho ue to the close spectral signature of species like the re cear an the fir for instance [8]. In [4], a classification ase on reflectance is use to separate the conifers from the eciuous trees. The internal structure an shaing within a crown offer other ifferentiating criteria. One such measure, the proportion of re an clear pixels to the total numer of pixels, can ientify irch trees. Another helps ifferentiate aspens from spruces. A hierarchy of criteria is set forth to classify the crowns. Classification accuracy using this strategy was 75%. Raiometry an texture analysis have een use extensively in remote sensing applications. In this paper, we incorporate information otaine from stuying tree crown shapes to improve classification performance. As in the Erikson [4] stuy, the tree classification is performe on the four most prevalent species of in Sween: orway spruce, Scots pine, irch, an aspen. Two of these species are coniferous while the other two are eciuous. We select 48 crowns (12 per class) from the high resolution CIR images (Fig.1). Their contours, represente y a set of points, are then extracte in orer to stuy their shapes. The support vector machine (SVM), a supervise learning metho, was chosen for the classification. An important property of this classifier is that uring the training process, only a small suset of the training set vectors are selecte as support vectors. This reuces the computational cost an provies etter generalization, so that, for instance, when new samples far from the ecision ounary are introuce, the existing support vectors remain unchange. Figure 1: An example of one of the images use in this work (resolution 3 cm), c Sweish University of Agricultural Sciences.
2 2. SUPPORT VECTOR MACHIE w.x + = 1 Optimal Separating Hyperplane SVMs are a set of supervise learning methos of classification an regression, which consists in fining the maximum separation (margin) etween classes using a set of oservations calle the training ata. SVMs are also calle maximum margin classifiers since they minimize the empirical classification error an maximize the geometric margin simultaneously [15]. y i = 1 x l ξ l Support Vectors 2.1 Linear SVM Separale case Given a training ata (x i,y i ), i = 1,...,, where x i R m an y i { 1,1}, that enotes to which class x i elongs, SVM looks for the Optimal Separating Hyperplane that maximizes the istance etween the closest training samples of the two classes calle also the margin. The iviing hyperplane is efine as following w x + = 0. The vector w is a vector normal to the hyperplane an is the offset parameter allowing us to increase the margin. So the classifier is given y f : x R m sign(w x+) { 1,1}, i.e. all the training ata satisfy the following constraints: { w xi + 0 if y i = +1 w x i + 0 if y i = 1 That coul e escrie y a set of inequalities: y i (w x i + ) 1, i (1) To fin the hyperplane which gives the maximum margin 2/ w (for etails see [2]) we shoul minimize w 2, suject to constraints (1). This leas to the following quaratic optimization prolem: w 2 min (w,) 2 By introucing the Lagrange multipliers λ i,i = 1,..,, one for each inequality constraints (1), the prolem ecomes ual [2]: max (L(λ) = λ i λ 1 2 λ i λ j y i y j x i x j ), j=1 suject to λ i y i = 0, 0 λ i, i, which is a convex quaratic optimization prolem suject to linear constraints. Thus, the solution is given y w = λ iy i x i an = y i w x i, for i : λ i 0. The classification function ecomes: onseparale case f (x) = λ i y i x i x + For nonlinearly separale ata, slack variales ξ i an a regularization parameter C are introuce to eal with misclassifie samples, i.e. to relax the constraints (see Fig. 2). By introucing them into the optimization prolem, we otain: w 2 min (w,) 2 +C ξ i, suject to : y i (w x i + ) 1 ξ i, ξ i 0, i. x i margin = 2 w ξ k x k w.x + = 1 Figure 2: SVM Classifier. y j = 1 This quaratic prolem using the Lagrange multipliers ecomes: max (L(λ) = λ i λ 1 2 λ i λ j y i y j x i x j ), j=1 2.2 onlinear SVM suject to λ i y i = 0, 0 λ i C, i. For the cases where the ecision function is not a linear function of ata, the nonlinear classifier was create y applying the kernel trick (originally propose y Aizerman) to maximum-margin hyperplanes (cf [1]). The resulting algorithm is formally similar, except that every ot prouct x i x j is replace y a non-linear kernel function K(x i,x j ) = Φ(x i ) Φ(x j ), with Φ : R m H an H eing an Eucliean space generally of a higher imension. This allows the algorithm to fit the maximum-margin linear hyperplane in the transforme H. Thus, the classification function ecomes: f (x) = λ i y i K(x i,x) +. In regar to kernels, there exists a mapping Φ, if K(x i,x j ) satisfy the Mercer s conition [16]. 3. FEATURE CREATIO AD CLASSIFICATIO RESULTS As mentione in the introuction, 48 tree crown contours were selecte. The contours were carefully elineate manually to preserve important tree crown shape information. The SVM classification is performe using a Gaussian kernel K(x,x ) = exp( x x 2 ). Since 2σ 2 the ataase is quite small, for each experimental run, 50% of the samples were picke at ranom to form the training set. First, we performe the classification using only raiometric attriutes. We then a texture features an finally, we inclue the shape escriptors. After each step, we calculate the performance. To evaluate the performance, the average performance, P, of a set of experiments is compute. 5% of the values at the high an low en of the performance scale are exclue from this calculation. P can e expresse as e P = ( w P i P w P )/( e w ). w=1 =1
3 P i = c / is the performance of experiment, e the numer of experiments, c the numer of correctly classifie trees, the numer of trees an w an the numers of the worst an the est results respectively. The maximum performance P max is also given (from which we can get the est comination for training an test sets). Then, we present the confusion matrix for the est performance run. A confusion matrix is a visual representation of actual versus preicte classifications. 3.1 Raiometry ase features For vegetation an lan use monitoring, CIR or false-colour film offers a richer set of information than natural colour film. The false colours that escrie the three channels are Re(IR), Green(Re), Blue(Green), [12]. CIR photography is commonly use to iscriminate live healthy trees from ying vegetation. In our stuy, CIR allows us to istinguish the conifers from eciuous trees. As was pointe out in [4], the 4 classes of trees - aspen, irch, spruce an pine - can e easily ientifie as eciuous or coniferous from first orer statistics (the mean an stanar eviation, compute from the histogram of pixel intensities on the image). This is ecause eciuous trees reflect a sustantially greater percentage of infrare light. Classification ase exclusively on these escriptors gives low performance with the average performance eing P = 0.54 an the maximum, P max = The confusion matrix for one of the est values of P after repeate classification runs is shown elow a c where (a)aspen, ()irch, (c)spruce, ()pine. From this matrix, we can see that there are a sustantial numer of false positives. For example, 50% of aspens are classifie as irch, while 33% of the irch trees are classifie as either aspen or spruce. This classifier is especially weak in ifferentiating etween coniferous an eciuous classes, i.e. etween aspen an irch or etween spruce an pine. 3.2 Texture ase features To further istinguish within the eciuous an coniferous classes, we performe texture analysis using gray level co-occurrence matrices (GLCM), which is well aapte for characterising microtextures. A co-occurrence matrix is a two-imensional quantitative representation of spatial relationship [9]. Let {I(x,y), 0 x 1, 0 y 1} enote a image with G gray levels as escrie in [3]. The G G gray level co-occurrence matrix P is efine as is a measure of the homogeneity of the texture. If the gray level transitions are roughly uniformly istriute, the energy has a low value. Conversely, textures which have ominant gray level transition moes have higher energy values. The contrast feature i (i j) 2 P (i, j) j is a measure of local variation present in an image. We selecte this feature to exploit the istinctive features present in the four types of crown surface. Spruce trees, for example, isplay a raial pattern while aspen have ranom light an ark regions. Since these two features are inepenent of the size an shape of the crown surface, they represent pure texture characteristics. The evaluation of classifier performance allows us to etermine the optimal couple of parameters ( an irection). In practice we have chosen = 1 an 135 egree irection. By incorporating these two texture features into the classifier, we were ale to separate the trees into 4 classes with the average performance P = 0.71 an the maximum performance P max = The confusion matrix for one the est values of P after repeate classification runs is shown elow: We can see that the texture information allows to improve the classification results for esciuous, especially for aspens (a). 3.3 Shape ase features To further improve the previous results, we propose to stuy tree crown shapes in a so calle shape space, using the representation of planar shapes y their angle functions, cf [11]. We consier the tree crown contours as continuous an close curves in R 2. The representation of such curves in the shape space is invariant to rigi rotation an translation, an to scaling in R 2. Let us efine these properties. Curves α = (α 1 (s),α 2 (s)) are parametrize y arclength s, where α : R R 2 with perio 2π satisfying conition of constant spee along the curve α (s) = 1, s. We can write α (s) = e jθ(s), where θ : R R an j = 1 associating C with R 2. θ is calle irection function or angle function. For every s, θ(s) gives the angle etween α (s) an the positive ascissa x, cf Fig.3: a c P (i, j) = ((r,s),(t,v)) : I(r,s) = i,i(t,v) = j. The entry (i, j) of matrix P is the numer of occurrences of the pair of gray levels i an j which are a istance = (x,y) apart. Since is a isplacement vector, (r,s), (t,v), such that (t,v) = (r + x, s + y), an. is the carinality of a set. GLCMs are a compact representation of pairs of pixel values in relation to each other. They are an example of the secon orer statistics as efine y Julesz an several useful features can e compute from them. We generate 9 such matrices for each tree, each matrix representing a ifferent irection or istance. Two texture features, energy an contrast were extracte from the GLCMs. The energy term, given y i P 2 (i, j) j Figure 3: Examples of shapes an corresponing angle functions. On the Fig.4 we can see more complicate graphs of angle function representing some crowns.
4 Figure 4: Examples of crowns of four species, their contours, corresponing angle functions θ, an θ = θ θ 0 at the ottom. (a)aspen, ()Birch, (c)spruce, ()Pine. We assume L 2 to enote the space of real functions R R of the perio 2π an square integrale on [0,2π] with scalar prouct f 1, f 2 = 0 f 1 (s) f 2 (s)s an f = f, f. Let θ(s) e the angle function of a planar shape. For the unit circle the function is θ 0 (s) = s. For any other close curve the angle function can e written as θ = θ 0 + f, where f L 2. The space θ 0 + L 2 is an affine space the elements of which are ifferent y an element of L 2. The properties of curve invariance mentione aove are guarantee y the following conitions (cf [11]): 1. The prolem of scaling can e simply resolve y fixing the length of the curve, for instance y 2π. Let c e a close contour, then we otain: s = s = 2π; c 0 2. An aition of a constant to the angle function θ is equivalent to the rotation of the curve in R 2. To guarantee the invariance of the curve to this action, we eal with the functions θ, the mean values of which are equal to a constant on [0,2π]. So, let 1 θ(s)s = π, where constant π is chosen to inclue the ientity function θ 0 in the restricte set, since 1 2π θ 0 (s)s = 1 ss = π; 3. Finally, to e close, the curves must satisfy following conition: exp( jθ(s))s = 0. 0 We efine C θ 0 + L 2 as the set of all the elements of θ 0 + L 2 satisfying the conitions 1-3 escrie aove. Or more formally, efine the map φ = (φ 1,φ 2,φ 3 )) : (θ 0 + L 2 ) R 3 : φ 1 = 1 θ(s)s φ 2 = 1 cos(θ(s))s φ 3 = 1 sin(θ(s))s. From now on, we can efine C as C = φ 1 (π,0,0), which is calle the pre-shape space, ecause the same curve with ifferent initial points (s = 0) correspon to ifferent elements of the space C. But we will eal with the shape space S, where such curves are represente y one element of S efine y C/D, where D is the unite circle. We are going to escrie the tree features create using such a representation of the crown contours. To o it some tree crown shape analysis was one to etermine the information that coul allow us to recognize the species. Let us look at Fig.4. We can see that the aspens have an irregular structure, the convexities/ranches sticking out of the oy of crowns. The ranches of spruces are more regular an have a more raial irection. The irch an pine crown contours are more roun, where irches are the smoothest. The first feature, chosen to e incorporate in the classifier, is a istance to a circle (θ,θ 0 ), which is compute using the geoesic on the shape space (see [11] for etails). For the given shapes θ i,i = 1,...,M, we perform a vector of istances to a circle : v c (θ) = {(θ i,θ 0 ), i = 1,2,...,M} R, where M is the numer of trees. We also inclue geometrical escriptors ase on θ = θ θ 0. Firstly, we translate the property that some species have more regular structure of the crowns than others, y a measure of contour elasticity: v e ( θ) = θ(s) 2 s, 0 v e ( θ) = {v e ( θ i ), i = 1,2,...,M} R. In regar to spruces, generally they have ig ranches/convexities an not numerous in comparison with the convexities of irch crown. This criterion is reflecte y the angle function uner the form of the local maxima numer. v ( θ) =, v ( θ) = {v ( θ i ), i = 1,2,...,M} R. Then, the pine crown shape is quite close to that of irch ut certain roughnesses are a little it igger, some ranches sticking out. Thus we can qualify the crown contour irregularities ue to the ranches, leaves an shaows: v µ ( θ) = µ = 1 n n θ(s k ), k=1 v µ ( θ) = {µ i, i = 1,2,...,M} R, where s k = k t=1 c t, c t = p t+1 p t chor lengths of contour an p t R 2,t = 1,...,n orere collection of points approximating the contours, v Var ( θ) = Var = 1 n n 1 (µ i θ(s k ) ) 2, k=1 v Var ( θ) = {Var i, i = 1,2,...,M} R. Finally, we use the ensity function of the angle function in orer to have information aout the quantity of ig an small convexities of the crown contour. For that, we calculate the histograms of θ, h θ (x) = car{ θ (s) = x,s [0,2π],x R}, an then Eucliean istances (h 1 (x),h 2 (x)) = < h 1 (x),h 2 (x) >. Then we create the following features associate to each shape, one istance for every class l. Such a istance is calculate as the istance to the nearest element of the class l: l min (h i) = min{(h i,h j )} j=1,... m l R,
5 v ( θ) = { l min (h i), i = 1,2,...,M, l = 1,..., l } R, where l is the numer of classes an m l is the numer of shapes of each class in the training set. ow we are forming a long vector using iniviual features an apply SVM on that feature vector. Otaine results: the average performance P = an the maximum of performances P max = An example of confusion matrix is given as follows: Shape ase features allow us to improve the ientification of species within the conifers an eciuous trees. 4. COCLUSIO In this paper we show that the performance of a classifier ase on conventional spectral an texture characteristics can e improve y incluing shape escriptors in the feature set. By incorporating shape features, the classification performance mean improve y aout 4% for 6 samples per class while the maximum performance was 87.5%. To create the new escriptors, the shapes of crowns were analyse using the angle function representation. Such representation allowe us to preserve the tree crown characteristics that associate it with one class or another. The limits impose y the ataase size i not give us much freeom to experiment with the training to test sample size ratio. We i, however, oserve that the performance mean tene to increase when samples were ae to the training set (see fig. 5). These results are encouraging ut they nee to e valiate on a larger numer of images an tree crowns. Figure 5: Performance means an maxima. Future work will consist in automatic extraction of tree crowns using shape information. a c REFERECES [1] B. E. Boser, I. M. Guyon an V.. Vapnik, A Training Algorithm for Optimal Margin Classifiers, in Fifth Annual Workshop on Computational Learning Theory, Pittsurg, [2] C. J. C. Burges, Tutorial on Support Vector Machines for Pattern recognition. Kluwer Acaemic Pulishers, Boston. pp [3] C. H. Chen, L. F. Pau, P. S. P. Wang, The Hanook of Pattern Recognition an Computer Vision (2n Eition), pp , Worl Scientific Pulishing Co., [4] M. Erikson, Segmentation an Classification of Iniviual Tree Crowns, Ph.D. thesis, Sweish University of Agricultural Sciences, Uppsala, Sween, [5] M. Erikson, Species Classification of Iniviually Segmente Tree Crowns in High-Resolution Aerial Images using Raiometric an Morphologic Image Measures, Remote Sensing of Environment, vol. 91, pp , April [6] F. A. Gougeon, Comparison of Possile Multispectral Classification Scheme for Tree Crown Iniviually Delineate on High Spatial Resolution MEIS Images, Canaian Journal of Remote Sensing, 21(1), pp. 1 9, [7] F. A. Gougeon, Vers l inventaire forestier automatisé : reconnaître l arre ou la forêt?, CD-ROM u 9 eme Congrès e L Association quéécoise e téléétection, May [8] F. A. Gougeon, D. G. Leckie, I. Scott an D.Paaraine Iniviual Tree Crown Species Recognition: the ahmint Stuy, Proc. of the International Forum on Automate Interpretation of High Spatial Resolution Digital Imagery for Forestry, pp , Pacific Forestry Center, Victoria, British Columia, Canaa, Feruary [9] R. M. Haralick, Statistical an Structural Approaches to Texture, Proceeings of the IEEE, vol. 67, pp , [10] D. G. Leckie, F. A. Gougeon,. Walsworth an D. Paraine, Stan elineation an Composition estimation using Semi- Automate Iniviual Tree Crown Analysis, Remote Sensing of Environment, vol. 85, pp , [11] E. Klassen, A. Srivastava, W. Mio an S. H. Joshi, Analysis of Planar Shapes Using Geoesic Paths on Shape Space, IEEE Transaction on Pattern Analysis an Machine Intelligence, vol. 26, no 3, pp , March [12] G. Perrin, Etue u Couvert Forestier par Processus Ponctuels Marqués, Ph.D. thesis, Ecole Doctoral e Centrale Paris, Octoer [13] G. Perrin, X. Descomes an J. Zeruia. A on-bayesian Moel for Tree Crown Extraction using Marke Point Processes, Research Report 5846, IRIA, France, Feruary, [14] A. Srivastava, S. H. Joshi, W. Mio an X. Liu, Statistical Shape Analysis: Clustering, Learning, an Testing, IEEE Transaction on Pattern Analysis an Machine Intellegence, vol. 27, no 4, pp , March [15] V.. Vapnik, Statistical Learning Theory. John Wily an sons, inc., [16] V.. Vapnik, The ature of Statistical Learning Theory. Springer Verlag, ew-york, 1995.
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 informationTHE 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 informationObject Recognition Using Colour, Shape and Affine Invariant Ratios
Object Recognition Using Colour, Shape an Affine Invariant Ratios Paul A. Walcott Centre for Information Engineering City University, Lonon EC1V 0HB, Englan P.A.Walcott@city.ac.uk Abstract This paper escribes
More informationClassifying 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 information1 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 informationImage 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 informationNEW METHOD FOR FINDING A REFERENCE POINT IN FINGERPRINT IMAGES WITH THE USE OF THE IPAN99 ALGORITHM 1. INTRODUCTION 2.
JOURNAL OF MEDICAL INFORMATICS & TECHNOLOGIES Vol. 13/009, ISSN 164-6037 Krzysztof WRÓBEL, Rafał DOROZ * fingerprint, reference point, IPAN99 NEW METHOD FOR FINDING A REFERENCE POINT IN FINGERPRINT IMAGES
More informationCONSTRUCTION AND ANALYSIS OF INVERSIONS IN S 2 AND H 2. Arunima Ray. Final Paper, MATH 399. Spring 2008 ABSTRACT
CONSTUCTION AN ANALYSIS OF INVESIONS IN S AN H Arunima ay Final Paper, MATH 399 Spring 008 ASTACT The construction use to otain inversions in two-imensional Eucliean space was moifie an applie to otain
More informationParticle 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 informationShift-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 informationExercises 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 informationResearch Article Research on Law s Mask Texture Analysis System Reliability
Research Journal of Applie Sciences, Engineering an Technology 7(19): 4002-4007, 2014 DOI:10.19026/rjaset.7.761 ISSN: 2040-7459; e-issn: 2040-7467 2014 Maxwell Scientific Publication Corp. Submitte: November
More informationFeature 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 informationOnline 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 informationRough 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 informationA COMPARATIVE STUDY OF THREE METHODS FOR IDENTIFYING INDIVIDUAL TREE CROWNS IN AERIAL IMAGES COVERING DIFFERENT TYPES OF FORESTS
A COMPARATIVE STUDY OF THREE METHODS FOR IDENTIFYING INDIVIDUAL TREE CROWNS IN AERIAL IMAGES COVERING DIFFERENT TYPES OF FORESTS Mats Eriksson, Guillaume Perrin, Xavier Descombes, Josiane Zerubia Ariana
More informationFigure 1: 2D arm. Figure 2: 2D arm with labelled angles
2D Kinematics Consier a robotic arm. We can sen it commans like, move that joint so it bens at an angle θ. Once we ve set each joint, that s all well an goo. More interesting, though, is the question of
More informationKinematic Analysis of a Family of 3R Manipulators
Kinematic Analysis of a Family of R Manipulators Maher Baili, Philippe Wenger an Damien Chablat Institut e Recherche en Communications et Cybernétique e Nantes, UMR C.N.R.S. 6597 1, rue e la Noë, BP 92101,
More informationUsing 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 informationQueueing Model and Optimization of Packet Dropping in Real-Time Wireless Sensor Networks
Queueing Moel an Optimization of Packet Dropping in Real-Time Wireless Sensor Networks Marc Aoun, Antonios Argyriou, Philips Research, Einhoven, 66AE, The Netherlans Department of Computer an Communication
More informationResearch 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 informationShift-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 informationPedestrian Detection Using Structured SVM
Pedestrian Detection Using Structured SVM Wonhui Kim Stanford University Department of Electrical Engineering wonhui@stanford.edu Seungmin Lee Stanford University Department of Electrical Engineering smlee729@stanford.edu.
More informationClassical Mechanics Examples (Lagrange Multipliers)
Classical Mechanics Examples (Lagrange Multipliers) Dipan Kumar Ghosh Physics Department, Inian Institute of Technology Bombay Powai, Mumbai 400076 September 3, 015 1 Introuction We have seen that the
More information5th International Conference on Advanced Design and Manufacturing Engineering (ICADME 2015)
5th International Conference on Avance Design an Manufacturing Engineering (ICADME 25) Research on motion characteristics an application of multi egree of freeom mechanism base on R-W metho Xiao-guang
More informationHW2 due on Thursday. Face Recognition: Dimensionality Reduction. Biometrics CSE 190 Lecture 11. Perceptron Revisited: Linear Separators
HW due on Thursday Face Recognition: Dimensionality Reduction Biometrics CSE 190 Lecture 11 CSE190, Winter 010 CSE190, Winter 010 Perceptron Revisited: Linear Separators Binary classification can be viewed
More informationA 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 informationLoop Scheduling and Partitions for Hiding Memory Latencies
Loop Scheuling an Partitions for Hiing Memory Latencies Fei Chen Ewin Hsing-Mean Sha Dept. of Computer Science an Engineering University of Notre Dame Notre Dame, IN 46556 Email: fchen,esha @cse.n.eu Tel:
More information6 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 informationRobot Learning. There are generally three types of robot learning: Learning from data. Learning by demonstration. Reinforcement learning
Robot Learning 1 General Pipeline 1. Data acquisition (e.g., from 3D sensors) 2. Feature extraction and representation construction 3. Robot learning: e.g., classification (recognition) or clustering (knowledge
More informationCharacterizing 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 informationNon-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 informationMANJUSHA K.*, ANAND KUMAR M., SOMAN K. P.
Journal of Engineering Science an echnology Vol. 13, No. 1 (2018) 141-157 School of Engineering, aylor s University IMPLEMENAION OF REJECION SRAEGIES INSIDE MALAYALAM CHARACER RECOGNIION SYSEM BASED ON
More informationBayesian 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 informationA 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 informationData Analysis 3. Support Vector Machines. Jan Platoš October 30, 2017
Data Analysis 3 Support Vector Machines Jan Platoš October 30, 2017 Department of Computer Science Faculty of Electrical Engineering and Computer Science VŠB - Technical University of Ostrava Table of
More informationA Classification of 3R Orthogonal Manipulators by the Topology of their Workspace
A Classification of R Orthogonal Manipulators by the Topology of their Workspace Maher aili, Philippe Wenger an Damien Chablat Institut e Recherche en Communications et Cybernétique e Nantes, UMR C.N.R.S.
More informationRobust 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 informationThreshold Based Data Aggregation Algorithm To Detect Rainfall Induced Landslides
Threshol Base Data Aggregation Algorithm To Detect Rainfall Inuce Lanslies Maneesha V. Ramesh P. V. Ushakumari Department of Computer Science Department of Mathematics Amrita School of Engineering Amrita
More informationSkyline 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 informationGeneralized Edge Coloring for Channel Assignment in Wireless Networks
Generalize Ege Coloring for Channel Assignment in Wireless Networks Chun-Chen Hsu Institute of Information Science Acaemia Sinica Taipei, Taiwan Da-wei Wang Jan-Jan Wu Institute of Information Science
More informationStudy 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 informationCAMERAS 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 informationArnold W.M Smeulders Theo Gevers. University of Amsterdam smeulders}
Arnold W.M Smeulders Theo evers University of Amsterdam email: smeulders@wins.uva.nl http://carol.wins.uva.nl/~{gevers smeulders} 0 Prolem statement Query matching Query 0 Prolem statement Query classes
More informationVision-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 informationGeneralized 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 informationAn ECA-based Control-rule formalism for the BPEL Process Modularization *
Availale online at www.scienceirect.com Proceia Environmental Sciences 11 (2011) 511 517 An ECA-ase Control-rule formalism for the BPEL Process Moularization * Bang Ouyang, Farong Zhong **, Huan Liu Department
More informationa 1 (x ) a 1 (x ) a 1 (x ) ) a 2 3 (x Input Variable x
Support Vector Learning for Fuzzy Rule-Base Classification Systems Yixin Chen, Stuent Member, IEEE, James Z. Wang, Member, IEEE Abstract To esign a fuzzy rule-base classification system (fuzzy classifier)
More informationSupport Vector Machines
Support Vector Machines RBF-networks Support Vector Machines Good Decision Boundary Optimization Problem Soft margin Hyperplane Non-linear Decision Boundary Kernel-Trick Approximation Accurancy Overtraining
More informationShort-term prediction of photovoltaic power based on GWPA - BP neural network model
Short-term preiction of photovoltaic power base on GWPA - BP neural networ moel Jian Di an Shanshan Meng School of orth China Electric Power University, Baoing. China Abstract In recent years, ue to China's
More informationI. 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 informationDEVELOPMENT OF DamageCALC APPLICATION FOR AUTOMATIC CALCULATION OF THE DAMAGE INDICATOR
Mechanical Testing an Diagnosis ISSN 2247 9635, 2012 (II), Volume 4, 28-36 DEVELOPMENT OF DamageCALC APPLICATION FOR AUTOMATIC CALCULATION OF THE DAMAGE INDICATOR Valentina GOLUBOVIĆ-BUGARSKI, Branislav
More informationSolution 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 informationArchitecture Design of Mobile Access Coordinated Wireless Sensor Networks
Architecture Design of Mobile Access Coorinate Wireless Sensor Networks Mai Abelhakim 1 Leonar E. Lightfoot Jian Ren 1 Tongtong Li 1 1 Department of Electrical & Computer Engineering, Michigan State University,
More informationA 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 informationA 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 informationFast 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 informationEXTRACTION 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 informationGenerative and discriminative classification techniques
Generative and discriminative classification techniques Machine Learning and Category Representation 013-014 Jakob Verbeek, December 13+0, 013 Course website: http://lear.inrialpes.fr/~verbeek/mlcr.13.14
More informationLearning 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 informationA 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 informationScale-Invariance of Support Vector Machines based on the Triangular Kernel. Abstract
Scale-Invariance of Support Vector Machines based on the Triangular Kernel François Fleuret Hichem Sahbi IMEDIA Research Group INRIA Domaine de Voluceau 78150 Le Chesnay, France Abstract This paper focuses
More informationModule13: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 informationSTATISTICAL TEXTURE ANALYSIS OF ROAD FOR MOVING OBJECTIVES
U.P.B. Sci. Bull., Series C, Vol. 70, Iss. 4, 008 ISSN 454-34x STATISTICA TEXTURE ANAYSIS OF ROAD FOR MOVING OBJECTIVES Dan POPESCU, Rau DOBRESCU, Nicoleta ANGEESCU 3 Obiectivul articolului este acela
More informationTexture 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 informationState 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 informationNon-Bayesian Classifiers Part II: Linear Discriminants and Support Vector Machines
Non-Bayesian Classifiers Part II: Linear Discriminants and Support Vector Machines Selim Aksoy Department of Computer Engineering Bilkent University saksoy@cs.bilkent.edu.tr CS 551, Spring 2007 c 2007,
More informationSUPPORT VECTOR MACHINES
SUPPORT VECTOR MACHINES Today Reading AIMA 18.9 Goals (Naïve Bayes classifiers) Support vector machines 1 Support Vector Machines (SVMs) SVMs are probably the most popular off-the-shelf classifier! Software
More informationCLASS BASED RATIOING EFFECT ON SUB-PIXEL SINGLE LAND COVER AUTOMATIC MAPPING
CLSS BSED RTIOIG EFFECT O SUB-PIXEL SIGLE LD COVER UTOMTIC MPPIG nil Kumar *a, Suresh Saggar b, a Inian Institute of Remote Sensing, Dehraun, anil@iirs.gov.in b Birla Institute of Technology & Science,
More informationAn 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 informationA 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 informationLearning 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 informationImage compression predicated on recurrent iterated function systems
2n International Conference on Mathematics & Statistics 16-19 June, 2008, Athens, Greece Image compression preicate on recurrent iterate function systems Chol-Hui Yun *, Metzler W. a an Barski M. a * Faculty
More informationLecture 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 informationAn 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 informationHandling 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 informationOpen Research Online The Open University s repository of research publications and other research outputs
Open Research Online The Open University s repository of research publications an other research outputs Geometric configuration in robot kinematic esign Book Section How to cite: Rooney, Joe (006). Geometric
More informationCMSC 430 Introduction to Compilers. Spring Register Allocation
CMSC 430 Introuction to Compilers Spring 2016 Register Allocation Introuction Change coe that uses an unoune set of virtual registers to coe that uses a finite set of actual regs For ytecoe targets, can
More informationDual Arm Robot Research Report
Dual Arm Robot Research Report Analytical Inverse Kinematics Solution for Moularize Dual-Arm Robot With offset at shouler an wrist Motivation an Abstract Generally, an inustrial manipulator such as PUMA
More informationFast 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 informationAlmost 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 informationDiscriminative classifiers for image recognition
Discriminative classifiers for image recognition May 26 th, 2015 Yong Jae Lee UC Davis Outline Last time: window-based generic object detection basic pipeline face detection with boosting as case study
More informationMORA: 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 informationEstimating 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 informationFuzzy Clustering in Parallel Universes
Fuzzy Clustering in Parallel Universes Bern Wisweel an Michael R. Berthol ALTANA-Chair for Bioinformatics an Information Mining Department of Computer an Information Science, University of Konstanz 78457
More informationApproximation 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 informationTHE APPLICATION OF ARTICLE k-th SHORTEST TIME PATH ALGORITHM
International Journal of Physics an Mathematical Sciences ISSN: 2277-2111 (Online) 2016 Vol. 6 (1) January-March, pp. 24-6/Mao an Shi. THE APPLICATION OF ARTICLE k-th SHORTEST TIME PATH ALGORITHM Hua Mao
More informationA 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 informationEXACT 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 informationA 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 informationA Multi-scale Boosted Detector for Efficient and Robust Gesture Recognition
A Multi-scale Booste Detector for Efficient an Robust Gesture Recognition Camille Monnier, Stan German, Anrey Ost Charles River Analytics Cambrige, MA, USA Abstract. We present an approach to etecting
More information2-connected graphs with small 2-connected dominating sets
2-connecte graphs with small 2-connecte ominating sets Yair Caro, Raphael Yuster 1 Department of Mathematics, University of Haifa at Oranim, Tivon 36006, Israel Abstract Let G be a 2-connecte graph. A
More informationAdvanced 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 informationCONTENT-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 informationIMPROVING THE RELIABILITY OF DETECTION OF LSB REPLACEMENT STEGANOGRAPHY
IMPROVING THE RELIABILITY OF DETECTION OF LSB REPLACEMENT STEGANOGRAPHY Shreelekshmi R, Wilscy M 2 and C E Veni Madhavan 3 Department of Computer Science & Engineering, College of Engineering, Trivandrum,
More informationA Modified Immune Network Optimization Algorithm
IAENG International Journal of Computer Science, 4:4, IJCS_4_4_3 A Moifie Immune Network Optimization Algorithm Lu Hong, Joarer Kamruzzaman Abstract This stuy proposes a moifie artificial immune network
More informationThe Berlin Airlift ( ) was an operation by the United States and Great. Linear Programming. Preview Exercises
790 Chapter 7 Sstems of Equations an Inequalities 108. The reason that sstems of linear inequalities are appropriate for moeling health weight is ecause guielines give health weight ranges, rather than
More informationMulti-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 informationA 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 informationCluster 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