1. Introduction. Abstract

Size: px
Start display at page:

Download "1. Introduction. Abstract"

Transcription

1 Image Retreval Usng a Herarchy of Clusters Danela Stan & Ishwar K. Seth Intellgent Informaton Engneerng Laboratory, Department of Computer Scence & Engneerng, Oaland Unversty, Rochester, Mchgan dstan@oaland.edu & seth@oaland.edu Abstract The goal of ths paper s to descrbe an effcent procedure for color-based mage retreval. The proposed procedure conssts of two stages. Frst, the mage data set s herarchcally decomposed nto dsjont subsets by applyng an adaptaton of the -means clusterng algorthm. Snce Eucldean measure may not effectvely reproduce human percepton of a vsual content, the adaptve algorthm uses a non- Eucldean smlarty metrc and clustrods as cluster prototypes. Second, the derved herarchy s searched by a branch and bound method to facltate rapd calculaton of the -nearest neghbors for retreval n a raned order. The proposed procedure has the advantage of handlng hgh dmensonal data, and dealng wth non-eucldean smlarty metrcs n order to explore the nature of the mage feature vectors. The herarchy also provdes users wth a tool for quc browsng. 1. Introducton The ncreasng rate at whch mages are generated n many applcaton areas, gves rse to the need of mage retreval systems to provde an effectve and effcent access to mage databases, based on ther vsual content. Whle t s perfectly feasble to dentfy a desred mage from a small collecton smply by browsng, technques that are more effectve are needed wth collectons contanng thousands, or mllons of tems. The current mage retreval technques can be classfed accordng to the type and nature of the features used for ndexng and retreval. Keyword ndexng technques manually assgn eywords or classfcaton codes to each mage when t s frst added to the collecton and use these descrptors as retreval eys at search tme. Ther advantages consst of hgh expressve power, possblty to descrbe mage content from the level of prmtve features to the level of abstract features, nvolvng a sgnfcant amount of reasonng about the meanng and purpose of the objects or scenes depcted. On the other hand, manual ndexng presents few drawbacs regardng the usefulness of the assgned eywords and the ndexng tme. Snce the same pcture can have dfferent meanngs for dfferent people, dfferent eywords could be assocated wth the same pcture [1]. When the ndexng tme for every mage taes few mnutes, to ndex a collecton of mllon mages s an ntensve and tme consumng wor. Methods that permt mage searchng based on features automatcally extracted from the mages themselves are referred as content-based mage retreval (CBIR)

2 technques [2]. Color retreval yelds the best results, n that the computer results of color smlarty are smlar to those derved by a human vsual system [3]. The retreval becomes more effcent when the spatal arrangement and couplng of colors over the mage are taen nto account or when one more low-level feature, such as texture or shape, s added to the system. To be along wth the user s percepton of mage chromatc contents, the mages are parttoned nto blocs and a color hstogram s calculated for each bloc. In ths case, smlarty matchng also consders adjacent condtons among blocs wth smlar hstograms. The most notable example of queryng by color s IBM s QBIC system [4] that has been appled successfully n color matchng of tems n electronc mal order catalogues. One drawbac of the current content-based mage retreval systems s ther lmtaton to the low level features even f some researchers have attempted to fll the gap between low-level features and semantc features, by dervng hghlevel semantc concepts (harmony, dsharmony, calmness, exctement) from color arrangements [5]. Another problem wth the exstng mage retreval systems s that these systems do not provde a summary vew of the mages n ther database to ther users. The necessty of a summary vew appears when the user has no specfc query mage at the begnnng of the search process and wants to explore the mage collecton to locate mages of nterest [6]. The ndexng structure s also a bg ssue for CBIR systems. Image features are often very hgh dmensonal or the smlarty metrcs are too complex to have effcent ndexng structures. The exstng mult-dmensonal ndexng technques concentrate only on how to dentfy and mprove ndexng technques that are scalable to hgh dmensonal feature vectors n mage retreval [7]. The other nature of feature vectors n Image Retreval,.e. non-eucldean smlarty measures, cannot be explored usng structures that have been developed based on Eucldean dstance metrcs such as the -d trees, the R-d trees and ts varants. The goal of ths paper s to provde a CBIR system that s scalable to large sze mage collecton and s based on an effectve ndexng module that solves both hgh dmensonalty and non-eucldean nature of some color feature spaces. The module s bult usng an adaptaton of -means clusterng n whch the metrc s a non-eucldean smlarty metrc and the cluster prototype s desgned to summarze the cluster n a manner that s suted for quc human comprehenson of ts components. These prototypes gve the system the capablty of quc browsng through the entre mage collecton. The proposed system also uses a branch and bound tree-search module that appled to the herarchy of the resultant clusters wll facltate rapd calculaton of the nearest neghbors for retreval. The paper s organzed as follows. Secton 2 descrbes the color feature representaton of the mages from the database used n the proposed procedure. Secton 3 explans how the herarchy of smlar groups s bult by the adaptve - means algorthm and Secton 4 descrbes how the search s carred out by the branch and bound algorthm. Secton 5 consders the effectveness of the approach and how the user can browse elegantly through the mage database; these consderatons are expounded wth experments on a database of 2100 mages. The paper concludes wth some fnal comments and a note on future wor.

3 2. Color feature representaton Color s one of the most wdely used features for mage smlarty retreval. Ths s not surprsng gven the facts that color s an easly recognzable element of an mage and the human vsual system s capable of dfferentatng between nfntely large numbers of colors. In ths paper, we use the Color-WISE representaton for mage retreval descrbed n detal n [8]. The representaton s guded prmarly on three factors. Frst, the representaton must be closely related to human vsual percepton snce a user determnes whether a retreval operaton n response to an example query s successful or not. Color-WISE uses the HSV (hue, saturaton, value) color coordnate system that correlates well wth human color percepton and s commonly used by artsts to represent color nformaton present n mages. Second, the representaton must encode the spatal dstrbuton of color n an mage. Because of ths consderaton, Color-WISE system reles on a fxed parttonng scheme. Ths s n contrast wth several proposals n the lterature [9] suggestng color-based segmentaton to characterze the spatal dstrbuton of color nformaton. Although the color-based segmentaton approach provdes a more flexble representaton and hence more powerful queres, we beleve that these advantages are outweghed by the smplcty of the fxed parttonng approach. In the fxed parttonng scheme, each mage s dvded nto M N overlappng blocs as shown. The overlappng blocs allow a certan amount of fuzzy-ness to be ncorporated n the spatal dstrbuton of color nformaton, whch helps n obtanng a better performance. Three separate local hstograms (hue, saturaton and value) for each bloc are computed. The thrd factor consdered by the Color- WISE system s that fact that the representaton should be as compact as possble to mnmze storage and computaton efforts. To obtan a compact representaton, Color-Wse system extracts from each local hstogram the locaton of ts area-pea. Placng a fxed-szed wndow on the hstogram at every possble locaton, the hstogram area fallng wthn the wndow s calculated. The locaton of the wndow yeldng the hghest area determnes the hstogram area-pea. Ths value represents the correspondng hstogram. Thus, a more compact representaton s obtaned and each mage s reduced to 3 M N numbers (3 represents the number of hstograms for HSV). 3. Herarchy of clusters Clusterng s a dscovery process n data mnng. It groups a set of data n a way that maxmzes the smlarty wthn clusters and mnmzes the smlarty between two dfferent clusters. The dscovered clusters can explan the characterstcs of the underlyng data dstrbuton and serve as foundaton for other analyss technques [10]. Clusterng s also useful n mplementng the dvde and conquer strategy to reduce the computatonal complexty of varous decsonmang algorthms n pattern recognton.

4 We use a varaton of -means clusterng to buld a herarchy of clusters. At every level of the herarchy, the varaton of -means clusterng uses a non-eucldean smlarty metrc and the cluster prototype s desgned to summarze the cluster n a manner that s suted for quc human comprehenson of ts components. The resultant clusters are further dvded nto other dsjont sub-clusters performng organzaton of nformaton at several levels, gong for fner and fner dstnctons. The results of ths herarchy decomposton are represented by a tree structure n whch each node of the tree represents a cluster prototype and at the last level, each leaf represents an mage. The herarchy of the cluster prototypes gves the system the capablty of quc browsng through the entre mage collecton. Ths adaptaton of -means algorthm s requred snce the color trplets (hue, saturaton, and value) derved from RGB space by non-lnear transformaton, are not evenly dstrbuted n the HSV space; the representatve of a cluster calculated as a centrod also does not mae much sense n such a space. Instead of usng the Eucldean dstance, we need to defne a measure that s closer to the human percepton n the sense that the dstance between two color trplets s a better approxmaton to the dfference perceved by human. We present below the used smlarty metrc that taes nto account both the perceptual smlarty between the dfferent hstograms bns and the fact that human percepton s more senstve to changes n hue values; we also present how the cluster representatves are calculated and what s the splttng crteron. 3.1 Color smlarty metrc Clusterng methods requre that an ndex of proxmty or assocatons be establshed between pars of patterns [10]. A proxmty ndex s ether a smlarty or dssmlarty. The more two mages resemble each other, the larger a smlarty ndex and the smaller a dssmlarty ndex wll be. Snce our retreval system s desgned to retreve the most smlar mages wth a query mage, the proxmty ndex wll be defned wth respect to smlarty. Dfferent smlarty measures have been suggested n the lterature to compare mages [3, 11]. We are usng n our clusterng algorthm the smlarty measure that, besdes the perceptual smlarty between dfferent bns of a color hstogram, recognzes the fact that human percepton s more senstve to changes n hue values [8]. It also recognzes that human percepton s not proportonally senstve to changes n hue value. Let q and t represent the bloc number n a query Q and an mage T, respectvely. Let ( h q, s ) q, v q and ( h ) t, s t, v t represent the domnant huesaturaton par of the selected bloc n the query mage and n the mage T, respectvely. The bloc smlarty s defned by the followng relatonshp:

5 S ( q, t ) 1+ a D h ( h, h ) + b* D ( s, s ) + c D ( v, v ) q t 1 s q t v q t (1) Here D, D and h s D v represent the functons that measure smlarty n hue, saturaton and value. The constants a, b and c defne the relatve mportance of hue, saturaton and value n smlarty components. Snce human percepton s more senstve to hue, a hgher value s assgned to a than to b. The followng functon was used to calculate D : h D h ( h, h ) q t 1 cos h q h 2 t 2π 256 (2) The functon Dh explctly taes nto account the fact that hue s measured as an angle. Through emprcal evaluatons, a value of equal to two provdes a good non-lnearty n the smlarty measure to approxmate the subjectve judgment of the hue smlarty. The saturaton smlarty s calculated by: D ( s, s ) s q t s q s 256 t (3) The value smlarty s calculated by usng the same formula as for saturaton smlarty. Usng the smlartes between the correspondng blocs from the query Q and mage T, the smlarty between a query and an mage s calculated by the followng expresson: S( Q, T ) M N m S 1 M N 1 ( q, t ) The quantty m n the above expresson represents the masng bt for bloc and M N stands for the number of blocs Cluster prototypes The cluster prototypes are desgned to summarze the clusters n a manner that s suted for quc human comprehenson of ts components. They wll nform the user about the approxmate regon n whch clusters and ther descendants are m (4)

6 found. By buldng the herarchcal tree havng the cluster prototypes as nteror nodes, the system wll allow users to browse the mage collecton at dfferent levels of detals. We defne the cluster prototype to be the most smlar mage to the other mages from the correspondng cluster; n another words, the cluster representatve s the clustrod pont n the feature space,.e., the pont n the cluster that maxmzes the sum of the squares of the smlarty values to the other ponts of the cluster. If C s a cluster, ts clustrod M s expressed as: 2 M arg max S ( I, J ) I C J C S, s ther smlarty value. We use arg to denote that the clustrod s the argument (mage) for whch the maxmum of the sums s obtaned. Here I and J stand for any two mages from the cluster C and ( I J ) 3.3. Splttng crteron To buld a partton for a specfed number of clusters K, a splttng crteron s necessary to be defned. Snce the herarchy ams to support smlarty searches, we would le nearby feature vectors to be collected n the same or nearby nodes. Thus, the splttng crteron n our algorthm wll try to fnd an optmal partton (5) J e K 2 ( K ) w S ( I M ), 1 I C where, w 1 n that s defned as one that maxmzes the crteron sum-of-squared-error functon: M and I stand for the clustrod and any mage from cluster C, respectvely; 2 S ( I, M ) represents the squared of the smlarty value between I and M, and n represents the number of elements of cluster C. The reason of maxmzng the crteron functon comes from the fact that the proxmty ndex measures the smlarty; that s, the larger a smlarty ndex value s, the more two mages resemble one another. Once the partton s obtaned, n order to valdate the clusters,.e. whether or not the samples form one more cluster, several steps are nvolved. Frst, we defne the null hypothess and the alternatve hypothess as follows: H 0 : there are exactly K clusters for the n samples, and H A : the samples form one more cluster. Accordng to the Neyman-Pearson paradgm [12], a decson as to whether or not to reject H 0 n favor of H A s made based on a statstcs T ( n). The statstc s nothng else than the cluster valdty ndex that s senstve to the structure n the data: (6)

7 T ( n) J J e e ( K ) ( K +1) (7) To obtan an approxmate crtcal value for the statstc, that s the ndex s large enough to be unusual, we use a threshold that taes nto account that, for large n, J e ( K) and J e ( K +1) follow a normal dstrbuton. Followng these consderatons, we consder the threshold τ defned n [13] as: 2 τ 1 π d α π d n d (8) The rejecton regon for the null hypothess at the p-percent sgnfcance level s: T ( n) < τ (9) The parameter α n (8) s determned from the probablty p that the null hypothess H 0 s rejected when t s true and d s the sample sze. The last nequalty provdes us wth a test for decdng whether the splttng of a cluster s justfed. 4. The browsng and search strategy The sgnfcant feature of our scheme s the possblty of quc browsng of the mage set when no query mage s specfed. The user can browse frst the hghest level of the tree representng the herarchy and get summary vews of the entre mage collecton n the form of the prototypes of the clusters at that level. By traversng down the tree, the user gets fner and fner detals from one level to another. Usng an analogy wth the vew layers defned usng a herarchy of selforganzaton maps [6], we can consder the frst level of the tree as a global vew level of the entre mage collecton, the ntermedate levels as regonal levels and the last layer of the tree as a local layer gvng the most detaled summary vews for the mages. Each node from the last layer ponts to a group of smlar mages named mage layer. When a query mage s present, the second phase of our algorthm s nvolved. The search strategy mples a branch and bound algorthm n order to facltate rapd calculaton of the -nearest neghbors for retreval. We use the method defned n [14] whch tests the nodes of the tree by two smple stoppng rules that elmnates the necessty of calculatng many dstances. The frst rule s meant to elmnate from consderaton the node and ts correspondng group of samples f the dstance between the query and the node (clustrod) s greater than the sum between the current dstance to the nearest neghbor and the farthest dstance from the centrod to any sample from the cluster. The second rule reduces the number of calculatons

8 of dstances between the query and the samples of the node that survved to rule 1. If the dstance from the query to the clustrod s greater than the sum between the current dstance to the nearest neghbor and the dstance from the clustrod to a sample, do not calculate the dstance between the sample and the query anymore. To perform smlarty search, the color representaton of the query mage s frst matched at the frst layer to determne the most smlar cluster prototypes (nodes) that should be searched further. We elmnate from consderatons each node from frst layer for whch rule 1 s satsfed. The matchng s then repeated for the chldren of one of the nodes from the prevous layer that survved to rule 1, and so on untl the last layer s reached, whch brngs out a group of mages that can be the most smlar to the query mage. We do not need to compare each one of these mages wth the query mage snce rule 2 flters out the mages that not satsfy t. For the mages that fnally survve, the dstances to the query mage are calculated and ordered to fnd the current nearest neghbors. Then the algorthm s appled for the next node that was carred on after applyng rule 1 and the table of the current nearest neghbors s updated as needed. 5. Expermental Results We evaluate our algorthm for browsng and retreval on an mage database of 2100 mages. The color vector representaton of each mage has 3*8*8 elements snce each mage s parttoned nto 8*8 overlappng blocs and the mage color content s characterzed by three components: hue, saturaton and ntensty. To perform color-based smlarty retreval, the values of the constants ( a, b and c ) n formula (1) are expermentally chosen as beng 2.5, 0.5 and 0, respectvely. We rescale hue and saturaton to values between 0 and 255. In order to obtan the frst level (global layer) of the herarchy, we apply -means algorthm for 2, 3 and at each consequent, the cluster valdty s checed, to ensure that the number of elements n every cluster s a moderate one and the sum-of-squared-error crteron to be satsfed. Comparng the values of the test statstc (7) and the values of the threshold (8) wth respect to nequalty (9), the possble number of clusters for dfferent small values of the sgnfcance level s obtaned. Snce the value of the statstc for K 31 s greater than the threshold for consecutve small values of p, we choose the value of K to be 30. Further, we splt the nodes havng at least 30 mages (at least 2% out of the data set) by applyng -means algorthm agan and so, a lower level (regonal layer) of the herarchy s obtaned. The mnmum number of elements n every cluster to go further wth splttng s decded as a compromse between the sze of the termnal nodes and the number of nodes n the tree. Fewer elements n the fnal groups produce fewer dstance computatons n the retreval stage, but larger number of dstance computaton n the search stage. We end up wth a search tree havng 81 nodes, 4 levels and an average of 40 mages per termnal node. Fg. 2 shows a retreval result for browsng mode. The user browses the frst level (global layer) of the tree and hypothetcally speang, the user decdes to loo for mages smlar wth the prototype of cluster 9. The mage wll be updated wth the mages (found n herarchcal clusterng process) that are close to the centers of

9 clusters at the next layer. Assumng that the user decdes to see mages smlar wth the frst prototype of the second layer, the thrd layer (mage layer) wll dsplay the group of mages smlar wth the prevous chosen prototype. Global layer Local layer Image layer Fgure 2. Fg. 3 shows a retreval result for search mode. The mage query s n the top left of the mage. The user wants to retreve the most three smlar mages wth the mage query. Applyng the proposed scheme, the followng nodes are reached n order to fnd the 3-nearest neghbors: node 5 at frst level, node 28 at second layer, 76 at the fnal level. The nearest neghbors are pced up from the group of mages ponted by node 76. Query mage Fgure 3. For more results on color smlarty retreval vst our home page at 6. Conclusons and future wor Ths paper presented an effcent method for mage retreval. Snce the proposed procedure organzes the color nformaton as a herarchy of dfferent clusters, the user s provded wth summary vews of the entre mage collecton at dfferent level of detals. Fast calculaton of the -nearest neghbors s possble by usng a branch and bound algorthm as a search strategy. As future wor, we want to experment our system wth semantc features n addton to the low level ones. Snce browsng computerzed nformaton has a socal dmenson, we wll also

10 develop an nterface for better vsualzaton of the nformaton patterns beng browsed and more effectve means of communcatng the browsng process. References 1. Enser PGB. Pctoral nformaton retreval. Journal of Documentaton, 51(2), pp , Lew MS, Hujsmans DP, Denteneer D. Content based mage retreval: Optmal eys, Texture, Projectons or Templates. Image Databases and Mult-Meda Search, 1997; 8; Faloutsos C, Equtz W, Flcner M. Effcent and Effectve Queryng by Image Content. Journal of Intellgent Informaton Systems 3, 1994; Nblac W, Barber R, Equtz W. et al. The QBIC project: queryng mages by content usng color, texture, shape. Proc. SPIE: Storage and Retreval for Images and Vdeo Databases 1993; 1908; Corrdon J, Delbmbo A, Pala P. Image retreval by color semantcs. ACM Multmeda System Journal, Seth IK, Coman I. Image retreval usng herarchcal self-organzng feature maps. Pattern Recognton Letters 20, 1999; Ru Y, Huang TS, Chang SF. Image Retreval: Past, Present, and Future. Journal of Vsual Communcaton and Image Representaton, 1999; 10; Seth IK, Coman I, Day B et al. Color-WISE: A system for mage smlarty retreval usng color. Proc. SPIE: Storage and Retreval for Image and Vdeo Databases, 1998; 3312; Smth JR and Chang SF. Tools and Technques for Color Image Retreval. Proceedngs of the SPIE: Storage and Retreval for Image and Vdeo Databases IV, 1996; 2670; Jan AK, Dubes RC. Algorthms for Clusterng Data. Prentces Hall Advanced Reference Seres, Swan MJ, Ballard DH. Color Indexng. Internatonal Journal of Computer Vson, 1991; 7(1); Rce JA. Mathematcal Statstcs and Data Analyss. Duxbury Press, Duda RO, Hart PE. Pattern classfcaton and scene analyss. John Wley & Sons, Inc., Fuanaga K, Narendra PM. A branch and bound algorthm for computng -nearest neghbors. IEEE Transactons on Computers, 1975; C -24; Danela Stan s a graduate student n Computer Scence and Engneerng at Oaland Unversty. She receved her M.A. n Computer Scence n 1999 from Wayne State Unversty and B.S. n Mathematcs n 1993 from Unversty of Bucharest, Romana. She s currently nvolved n the area of content -based mage retreval and data mnng. Ishwar K. Seth s currently Professor and Char of Computer Scence and Engneerng at Oaland Unversty. Hs research nterests nclude pattern recognton, data mnng, and multmeda nformaton processng and ndexng. Professor Seth serves on the edtoral

11 boards of several nternatonal journals ncludng IEEE Trans. Pattern Analyss and Machne Intellgence and IEEE Multmeda.

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data A Fast Content-Based Multmeda Retreval Technque Usng Compressed Data Borko Furht and Pornvt Saksobhavvat NSF Multmeda Laboratory Florda Atlantc Unversty, Boca Raton, Florda 3343 ABSTRACT In ths paper,

More information

Content Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers

Content Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers IOSR Journal of Electroncs and Communcaton Engneerng (IOSR-JECE) e-issn: 78-834,p- ISSN: 78-8735.Volume 9, Issue, Ver. IV (Mar - Apr. 04), PP 0-07 Content Based Image Retreval Usng -D Dscrete Wavelet wth

More information

A Binarization Algorithm specialized on Document Images and Photos

A Binarization Algorithm specialized on Document Images and Photos A Bnarzaton Algorthm specalzed on Document mages and Photos Ergna Kavalleratou Dept. of nformaton and Communcaton Systems Engneerng Unversty of the Aegean kavalleratou@aegean.gr Abstract n ths paper, a

More information

Searching Large Image Databases using Color Information

Searching Large Image Databases using Color Information Searchng Large Image Databases usng Color Informaton Ioan Racu CMSC350: Artfcal Intellgence Wnter Quarter 2004 Department of Computer Scence Unversty of Chcago racu@cs.uchcago.edu ABSTRACT The goal of

More information

Cluster Analysis of Electrical Behavior

Cluster Analysis of Electrical Behavior Journal of Computer and Communcatons, 205, 3, 88-93 Publshed Onlne May 205 n ScRes. http://www.scrp.org/ournal/cc http://dx.do.org/0.4236/cc.205.350 Cluster Analyss of Electrcal Behavor Ln Lu Ln Lu, School

More information

Parallelism for Nested Loops with Non-uniform and Flow Dependences

Parallelism for Nested Loops with Non-uniform and Flow Dependences Parallelsm for Nested Loops wth Non-unform and Flow Dependences Sam-Jn Jeong Dept. of Informaton & Communcaton Engneerng, Cheonan Unversty, 5, Anseo-dong, Cheonan, Chungnam, 330-80, Korea. seong@cheonan.ac.kr

More information

An Image Fusion Approach Based on Segmentation Region

An Image Fusion Approach Based on Segmentation Region Rong Wang, L-Qun Gao, Shu Yang, Yu-Hua Cha, and Yan-Chun Lu An Image Fuson Approach Based On Segmentaton Regon An Image Fuson Approach Based on Segmentaton Regon Rong Wang, L-Qun Gao, Shu Yang 3, Yu-Hua

More information

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points; Subspace clusterng Clusterng Fundamental to all clusterng technques s the choce of dstance measure between data ponts; D q ( ) ( ) 2 x x = x x, j k = 1 k jk Squared Eucldean dstance Assumpton: All features

More information

MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION

MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION Paulo Quntlano 1 & Antono Santa-Rosa 1 Federal Polce Department, Brasla, Brazl. E-mals: quntlano.pqs@dpf.gov.br and

More information

Hierarchical clustering for gene expression data analysis

Hierarchical clustering for gene expression data analysis Herarchcal clusterng for gene expresson data analyss Gorgo Valentn e-mal: valentn@ds.unm.t Clusterng of Mcroarray Data. Clusterng of gene expresson profles (rows) => dscovery of co-regulated and functonally

More information

Background Removal in Image indexing and Retrieval

Background Removal in Image indexing and Retrieval Background Removal n Image ndexng and Retreval Y Lu and Hong Guo Department of Electrcal and Computer Engneerng The Unversty of Mchgan-Dearborn Dearborn Mchgan 4818-1491, U.S.A. Voce: 313-593-508, Fax:

More information

Fuzzy C-Means Initialized by Fixed Threshold Clustering for Improving Image Retrieval

Fuzzy C-Means Initialized by Fixed Threshold Clustering for Improving Image Retrieval Fuzzy -Means Intalzed by Fxed Threshold lusterng for Improvng Image Retreval NAWARA HANSIRI, SIRIPORN SUPRATID,HOM KIMPAN 3 Faculty of Informaton Technology Rangst Unversty Muang-Ake, Paholyotn Road, Patumtan,

More information

FEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur

FEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur FEATURE EXTRACTION Dr. K.Vjayarekha Assocate Dean School of Electrcal and Electroncs Engneerng SASTRA Unversty, Thanjavur613 41 Jont Intatve of IITs and IISc Funded by MHRD Page 1 of 8 Table of Contents

More information

Object-Based Techniques for Image Retrieval

Object-Based Techniques for Image Retrieval 54 Zhang, Gao, & Luo Chapter VII Object-Based Technques for Image Retreval Y. J. Zhang, Tsnghua Unversty, Chna Y. Y. Gao, Tsnghua Unversty, Chna Y. Luo, Tsnghua Unversty, Chna ABSTRACT To overcome the

More information

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching A Fast Vsual Trackng Algorthm Based on Crcle Pxels Matchng Zhqang Hou hou_zhq@sohu.com Chongzhao Han czhan@mal.xjtu.edu.cn Ln Zheng Abstract: A fast vsual trackng algorthm based on crcle pxels matchng

More information

A Clustering Algorithm for Key Frame Extraction Based on Density Peak

A Clustering Algorithm for Key Frame Extraction Based on Density Peak Journal of Computer and Communcatons, 2018, 6, 118-128 http://www.scrp.org/ournal/cc ISSN Onlne: 2327-5227 ISSN Prnt: 2327-5219 A Clusterng Algorthm for Key Frame Extracton Based on Densty Peak Hong Zhao

More information

A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION

A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION 1 THE PUBLISHING HOUSE PROCEEDINGS OF THE ROMANIAN ACADEMY, Seres A, OF THE ROMANIAN ACADEMY Volume 4, Number 2/2003, pp.000-000 A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION Tudor BARBU Insttute

More information

A Deflected Grid-based Algorithm for Clustering Analysis

A Deflected Grid-based Algorithm for Clustering Analysis A Deflected Grd-based Algorthm for Clusterng Analyss NANCY P. LIN, CHUNG-I CHANG, HAO-EN CHUEH, HUNG-JEN CHEN, WEI-HUA HAO Department of Computer Scence and Informaton Engneerng Tamkang Unversty 5 Yng-chuan

More information

Classifier Selection Based on Data Complexity Measures *

Classifier Selection Based on Data Complexity Measures * Classfer Selecton Based on Data Complexty Measures * Edth Hernández-Reyes, J.A. Carrasco-Ochoa, and J.Fco. Martínez-Trndad Natonal Insttute for Astrophyscs, Optcs and Electroncs, Lus Enrque Erro No.1 Sta.

More information

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

Learning the Kernel Parameters in Kernel Minimum Distance Classifier Learnng the Kernel Parameters n Kernel Mnmum Dstance Classfer Daoqang Zhang 1,, Songcan Chen and Zh-Hua Zhou 1* 1 Natonal Laboratory for Novel Software Technology Nanjng Unversty, Nanjng 193, Chna Department

More information

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices Steps for Computng the Dssmlarty, Entropy, Herfndahl-Hrschman and Accessblty (Gravty wth Competton) Indces I. Dssmlarty Index Measurement: The followng formula can be used to measure the evenness between

More information

Local Quaternary Patterns and Feature Local Quaternary Patterns

Local Quaternary Patterns and Feature Local Quaternary Patterns Local Quaternary Patterns and Feature Local Quaternary Patterns Jayu Gu and Chengjun Lu The Department of Computer Scence, New Jersey Insttute of Technology, Newark, NJ 0102, USA Abstract - Ths paper presents

More information

Unsupervised Learning and Clustering

Unsupervised Learning and Clustering Unsupervsed Learnng and Clusterng Why consder unlabeled samples?. Collectng and labelng large set of samples s costly Gettng recorded speech s free, labelng s tme consumng 2. Classfer could be desgned

More information

Edge Detection in Noisy Images Using the Support Vector Machines

Edge Detection in Noisy Images Using the Support Vector Machines Edge Detecton n Nosy Images Usng the Support Vector Machnes Hlaro Gómez-Moreno, Saturnno Maldonado-Bascón, Francsco López-Ferreras Sgnal Theory and Communcatons Department. Unversty of Alcalá Crta. Madrd-Barcelona

More information

Outline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1

Outline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1 4/14/011 Outlne Dscrmnatve classfers for mage recognton Wednesday, Aprl 13 Krsten Grauman UT-Austn Last tme: wndow-based generc obect detecton basc ppelne face detecton wth boostng as case study Today:

More information

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps Vsual Thesaurus for Color Image Retreval usng Self-Organzng Maps Chrstopher C. Yang and Mlo K. Yp Department of System Engneerng and Engneerng Management The Chnese Unversty of Hong Kong, Hong Kong ABSTRACT

More information

An Optimal Algorithm for Prufer Codes *

An Optimal Algorithm for Prufer Codes * J. Software Engneerng & Applcatons, 2009, 2: 111-115 do:10.4236/jsea.2009.22016 Publshed Onlne July 2009 (www.scrp.org/journal/jsea) An Optmal Algorthm for Prufer Codes * Xaodong Wang 1, 2, Le Wang 3,

More information

Outline. Type of Machine Learning. Examples of Application. Unsupervised Learning

Outline. Type of Machine Learning. Examples of Application. Unsupervised Learning Outlne Artfcal Intellgence and ts applcatons Lecture 8 Unsupervsed Learnng Professor Danel Yeung danyeung@eee.org Dr. Patrck Chan patrckchan@eee.org South Chna Unversty of Technology, Chna Introducton

More information

Machine Learning: Algorithms and Applications

Machine Learning: Algorithms and Applications 14/05/1 Machne Learnng: Algorthms and Applcatons Florano Zn Free Unversty of Bozen-Bolzano Faculty of Computer Scence Academc Year 011-01 Lecture 10: 14 May 01 Unsupervsed Learnng cont Sldes courtesy of

More information

COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL

COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL Nader Safavan and Shohreh Kasae Department of Computer Engneerng Sharf Unversty of Technology Tehran, Iran skasae@sharf.edu

More information

A Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems

A Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems A Unfed Framework for Semantcs and Feature Based Relevance Feedback n Image Retreval Systems Ye Lu *, Chunhu Hu 2, Xngquan Zhu 3*, HongJang Zhang 2, Qang Yang * School of Computng Scence Smon Fraser Unversty

More information

Semantic Image Retrieval Using Region Based Inverted File

Semantic Image Retrieval Using Region Based Inverted File Semantc Image Retreval Usng Regon Based Inverted Fle Dengsheng Zhang, Md Monrul Islam, Guoun Lu and Jn Hou 2 Gppsland School of Informaton Technology, Monash Unversty Churchll, VIC 3842, Australa E-mal:

More information

Query Clustering Using a Hybrid Query Similarity Measure

Query Clustering Using a Hybrid Query Similarity Measure Query clusterng usng a hybrd query smlarty measure Fu. L., Goh, D.H., & Foo, S. (2004). WSEAS Transacton on Computers, 3(3), 700-705. Query Clusterng Usng a Hybrd Query Smlarty Measure Ln Fu, Don Hoe-Lan

More information

An Improved Image Segmentation Algorithm Based on the Otsu Method

An Improved Image Segmentation Algorithm Based on the Otsu Method 3th ACIS Internatonal Conference on Software Engneerng, Artfcal Intellgence, Networkng arallel/dstrbuted Computng An Improved Image Segmentaton Algorthm Based on the Otsu Method Mengxng Huang, enjao Yu,

More information

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance Tsnghua Unversty at TAC 2009: Summarzng Mult-documents by Informaton Dstance Chong Long, Mnle Huang, Xaoyan Zhu State Key Laboratory of Intellgent Technology and Systems, Tsnghua Natonal Laboratory for

More information

Unsupervised Learning and Clustering

Unsupervised Learning and Clustering Unsupervsed Learnng and Clusterng Supervsed vs. Unsupervsed Learnng Up to now we consdered supervsed learnng scenaro, where we are gven 1. samples 1,, n 2. class labels for all samples 1,, n Ths s also

More information

Novel Pattern-based Fingerprint Recognition Technique Using 2D Wavelet Decomposition

Novel Pattern-based Fingerprint Recognition Technique Using 2D Wavelet Decomposition Mathematcal Methods for Informaton Scence and Economcs Novel Pattern-based Fngerprnt Recognton Technque Usng D Wavelet Decomposton TUDOR BARBU Insttute of Computer Scence of the Romanan Academy T. Codrescu,,

More information

Mathematics 256 a course in differential equations for engineering students

Mathematics 256 a course in differential equations for engineering students Mathematcs 56 a course n dfferental equatons for engneerng students Chapter 5. More effcent methods of numercal soluton Euler s method s qute neffcent. Because the error s essentally proportonal to the

More information

12/2/2009. Announcements. Parametric / Non-parametric. Case-Based Reasoning. Nearest-Neighbor on Images. Nearest-Neighbor Classification

12/2/2009. Announcements. Parametric / Non-parametric. Case-Based Reasoning. Nearest-Neighbor on Images. Nearest-Neighbor Classification Introducton to Artfcal Intellgence V22.0472-001 Fall 2009 Lecture 24: Nearest-Neghbors & Support Vector Machnes Rob Fergus Dept of Computer Scence, Courant Insttute, NYU Sldes from Danel Yeung, John DeNero

More information

Unsupervised Learning

Unsupervised Learning Pattern Recognton Lecture 8 Outlne Introducton Unsupervsed Learnng Parametrc VS Non-Parametrc Approach Mxture of Denstes Maxmum-Lkelhood Estmates Clusterng Prof. Danel Yeung School of Computer Scence and

More information

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique //00 :0 AM Outlne and Readng The Greedy Method The Greedy Method Technque (secton.) Fractonal Knapsack Problem (secton..) Task Schedulng (secton..) Mnmum Spannng Trees (secton.) Change Money Problem Greedy

More information

Machine Learning. Topic 6: Clustering

Machine Learning. Topic 6: Clustering Machne Learnng Topc 6: lusterng lusterng Groupng data nto (hopefully useful) sets. Thngs on the left Thngs on the rght Applcatons of lusterng Hypothess Generaton lusters mght suggest natural groups. Hypothess

More information

A Knowledge Management System for Organizing MEDLINE Database

A Knowledge Management System for Organizing MEDLINE Database A Knowledge Management System for Organzng MEDLINE Database Hyunk Km, Su-Shng Chen Computer and Informaton Scence Engneerng Department, Unversty of Florda, Ganesvlle, Florda 32611, USA Wth the exploson

More information

SCALABLE AND VISUALIZATION-ORIENTED CLUSTERING FOR EXPLORATORY SPATIAL ANALYSIS

SCALABLE AND VISUALIZATION-ORIENTED CLUSTERING FOR EXPLORATORY SPATIAL ANALYSIS SCALABLE AND VISUALIZATION-ORIENTED CLUSTERING FOR EXPLORATORY SPATIAL ANALYSIS J.H.Guan, F.B.Zhu, F.L.Ban a School of Computer, Spatal Informaton & Dgtal Engneerng Center, Wuhan Unversty, Wuhan, 430079,

More information

Detection of an Object by using Principal Component Analysis

Detection of an Object by using Principal Component Analysis Detecton of an Object by usng Prncpal Component Analyss 1. G. Nagaven, 2. Dr. T. Sreenvasulu Reddy 1. M.Tech, Department of EEE, SVUCE, Trupath, Inda. 2. Assoc. Professor, Department of ECE, SVUCE, Trupath,

More information

K-means and Hierarchical Clustering

K-means and Hierarchical Clustering Note to other teachers and users of these sldes. Andrew would be delghted f you found ths source materal useful n gvng your own lectures. Feel free to use these sldes verbatm, or to modfy them to ft your

More information

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS ARPN Journal of Engneerng and Appled Scences 006-017 Asan Research Publshng Network (ARPN). All rghts reserved. NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS Igor Grgoryev, Svetlana

More information

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration Improvement of Spatal Resoluton Usng BlockMatchng Based Moton Estmaton and Frame Integraton Danya Suga and Takayuk Hamamoto Graduate School of Engneerng, Tokyo Unversty of Scence, 6-3-1, Nuku, Katsuska-ku,

More information

Decision Strategies for Rating Objects in Knowledge-Shared Research Networks

Decision Strategies for Rating Objects in Knowledge-Shared Research Networks Decson Strateges for Ratng Objects n Knowledge-Shared Research etwors ALEXADRA GRACHAROVA *, HAS-JOACHM ER **, HASSA OUR ELD ** OM SUUROE ***, HARR ARAKSE *** * nsttute of Control and System Research,

More information

Related-Mode Attacks on CTR Encryption Mode

Related-Mode Attacks on CTR Encryption Mode Internatonal Journal of Network Securty, Vol.4, No.3, PP.282 287, May 2007 282 Related-Mode Attacks on CTR Encrypton Mode Dayn Wang, Dongda Ln, and Wenlng Wu (Correspondng author: Dayn Wang) Key Laboratory

More information

Load Balancing for Hex-Cell Interconnection Network

Load Balancing for Hex-Cell Interconnection Network Int. J. Communcatons, Network and System Scences,,, - Publshed Onlne Aprl n ScRes. http://www.scrp.org/journal/jcns http://dx.do.org/./jcns.. Load Balancng for Hex-Cell Interconnecton Network Saher Manaseer,

More information

Available online at Available online at Advanced in Control Engineering and Information Science

Available online at   Available online at   Advanced in Control Engineering and Information Science Avalable onlne at wwwscencedrectcom Avalable onlne at wwwscencedrectcom Proceda Proceda Engneerng Engneerng 00 (2011) 15000 000 (2011) 1642 1646 Proceda Engneerng wwwelsevercom/locate/proceda Advanced

More information

An Efficient Genetic Algorithm with Fuzzy c-means Clustering for Traveling Salesman Problem

An Efficient Genetic Algorithm with Fuzzy c-means Clustering for Traveling Salesman Problem An Effcent Genetc Algorthm wth Fuzzy c-means Clusterng for Travelng Salesman Problem Jong-Won Yoon and Sung-Bae Cho Dept. of Computer Scence Yonse Unversty Seoul, Korea jwyoon@sclab.yonse.ac.r, sbcho@cs.yonse.ac.r

More information

A new segmentation algorithm for medical volume image based on K-means clustering

A new segmentation algorithm for medical volume image based on K-means clustering Avalable onlne www.jocpr.com Journal of Chemcal and harmaceutcal Research, 2013, 5(12):113-117 Research Artcle ISSN : 0975-7384 CODEN(USA) : JCRC5 A new segmentaton algorthm for medcal volume mage based

More information

CS434a/541a: Pattern Recognition Prof. Olga Veksler. Lecture 15

CS434a/541a: Pattern Recognition Prof. Olga Veksler. Lecture 15 CS434a/541a: Pattern Recognton Prof. Olga Veksler Lecture 15 Today New Topc: Unsupervsed Learnng Supervsed vs. unsupervsed learnng Unsupervsed learnng Net Tme: parametrc unsupervsed learnng Today: nonparametrc

More information

Course Introduction. Algorithm 8/31/2017. COSC 320 Advanced Data Structures and Algorithms. COSC 320 Advanced Data Structures and Algorithms

Course Introduction. Algorithm 8/31/2017. COSC 320 Advanced Data Structures and Algorithms. COSC 320 Advanced Data Structures and Algorithms Course Introducton Course Topcs Exams, abs, Proects A quc loo at a few algorthms 1 Advanced Data Structures and Algorthms Descrpton: We are gong to dscuss algorthm complexty analyss, algorthm desgn technques

More information

Pictures at an Exhibition

Pictures at an Exhibition 1 Pctures at an Exhbton Stephane Kwan and Karen Zhu Department of Electrcal Engneerng Stanford Unversty, Stanford, CA 9405 Emal: {skwan1, kyzhu}@stanford.edu Abstract An mage processng algorthm s desgned

More information

A Novel Adaptive Descriptor Algorithm for Ternary Pattern Textures

A Novel Adaptive Descriptor Algorithm for Ternary Pattern Textures A Novel Adaptve Descrptor Algorthm for Ternary Pattern Textures Fahuan Hu 1,2, Guopng Lu 1 *, Zengwen Dong 1 1.School of Mechancal & Electrcal Engneerng, Nanchang Unversty, Nanchang, 330031, Chna; 2. School

More information

SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE

SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE Dorna Purcaru Faculty of Automaton, Computers and Electroncs Unersty of Craoa 13 Al. I. Cuza Street, Craoa RO-1100 ROMANIA E-mal: dpurcaru@electroncs.uc.ro

More information

Determining Fuzzy Sets for Quantitative Attributes in Data Mining Problems

Determining Fuzzy Sets for Quantitative Attributes in Data Mining Problems Determnng Fuzzy Sets for Quanttatve Attrbutes n Data Mnng Problems ATTILA GYENESEI Turku Centre for Computer Scence (TUCS) Unversty of Turku, Department of Computer Scence Lemmnkäsenkatu 4A, FIN-5 Turku

More information

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task Proceedngs of NTCIR-6 Workshop Meetng, May 15-18, 2007, Tokyo, Japan Term Weghtng Classfcaton System Usng the Ch-square Statstc for the Classfcaton Subtask at NTCIR-6 Patent Retreval Task Kotaro Hashmoto

More information

Sum of Linear and Fractional Multiobjective Programming Problem under Fuzzy Rules Constraints

Sum of Linear and Fractional Multiobjective Programming Problem under Fuzzy Rules Constraints Australan Journal of Basc and Appled Scences, 2(4): 1204-1208, 2008 ISSN 1991-8178 Sum of Lnear and Fractonal Multobjectve Programmng Problem under Fuzzy Rules Constrants 1 2 Sanjay Jan and Kalash Lachhwan

More information

Problem Set 3 Solutions

Problem Set 3 Solutions Introducton to Algorthms October 4, 2002 Massachusetts Insttute of Technology 6046J/18410J Professors Erk Demane and Shaf Goldwasser Handout 14 Problem Set 3 Solutons (Exercses were not to be turned n,

More information

A fast algorithm for color image segmentation

A fast algorithm for color image segmentation Unersty of Wollongong Research Onlne Faculty of Informatcs - Papers (Arche) Faculty of Engneerng and Informaton Scences 006 A fast algorthm for color mage segmentaton L. Dong Unersty of Wollongong, lju@uow.edu.au

More information

The Shortest Path of Touring Lines given in the Plane

The Shortest Path of Touring Lines given in the Plane Send Orders for Reprnts to reprnts@benthamscence.ae 262 The Open Cybernetcs & Systemcs Journal, 2015, 9, 262-267 The Shortest Path of Tourng Lnes gven n the Plane Open Access Ljuan Wang 1,2, Dandan He

More information

The Research of Support Vector Machine in Agricultural Data Classification

The Research of Support Vector Machine in Agricultural Data Classification The Research of Support Vector Machne n Agrcultural Data Classfcaton Le Sh, Qguo Duan, Xnmng Ma, Me Weng College of Informaton and Management Scence, HeNan Agrcultural Unversty, Zhengzhou 45000 Chna Zhengzhou

More information

A Multi-step Strategy for Shape Similarity Search In Kamon Image Database

A Multi-step Strategy for Shape Similarity Search In Kamon Image Database A Mult-step Strategy for Shape Smlarty Search In Kamon Image Database Paul W.H. Kwan, Kazuo Torach 2, Kesuke Kameyama 2, Junbn Gao 3, Nobuyuk Otsu 4 School of Mathematcs, Statstcs and Computer Scence,

More information

Brushlet Features for Texture Image Retrieval

Brushlet Features for Texture Image Retrieval DICTA00: Dgtal Image Computng Technques and Applcatons, 1 January 00, Melbourne, Australa 1 Brushlet Features for Texture Image Retreval Chbao Chen and Kap Luk Chan Informaton System Research Lab, School

More information

Reducing Frame Rate for Object Tracking

Reducing Frame Rate for Object Tracking Reducng Frame Rate for Object Trackng Pavel Korshunov 1 and We Tsang Oo 2 1 Natonal Unversty of Sngapore, Sngapore 11977, pavelkor@comp.nus.edu.sg 2 Natonal Unversty of Sngapore, Sngapore 11977, oowt@comp.nus.edu.sg

More information

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision SLAM Summer School 2006 Practcal 2: SLAM usng Monocular Vson Javer Cvera, Unversty of Zaragoza Andrew J. Davson, Imperal College London J.M.M Montel, Unversty of Zaragoza. josemar@unzar.es, jcvera@unzar.es,

More information

Mining User Similarity Using Spatial-temporal Intersection

Mining User Similarity Using Spatial-temporal Intersection www.ijcsi.org 215 Mnng User Smlarty Usng Spatal-temporal Intersecton Ymn Wang 1, Rumn Hu 1, Wenhua Huang 1 and Jun Chen 1 1 Natonal Engneerng Research Center for Multmeda Software, School of Computer,

More information

Efficient Mean Shift Algorithm based Color Images Categorization and Searching

Efficient Mean Shift Algorithm based Color Images Categorization and Searching 152 Effcent Mean Shft Algorthm based Color Images Categorzaton and Searchng 1 Dr S K Vay, 2 Sanay Rathore, 3 Abhshek Verma and 4 Hemra Sngh Thakur 1 Professor, Head of Dept Physcs, Govt Geetanal Grl s

More information

Skew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach

Skew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach Angle Estmaton and Correcton of Hand Wrtten, Textual and Large areas of Non-Textual Document Images: A Novel Approach D.R.Ramesh Babu Pyush M Kumat Mahesh D Dhannawat PES Insttute of Technology Research

More information

X- Chart Using ANOM Approach

X- Chart Using ANOM Approach ISSN 1684-8403 Journal of Statstcs Volume 17, 010, pp. 3-3 Abstract X- Chart Usng ANOM Approach Gullapall Chakravarth 1 and Chaluvad Venkateswara Rao Control lmts for ndvdual measurements (X) chart are

More information

A CALCULATION METHOD OF DEEP WEB ENTITIES RECOGNITION

A CALCULATION METHOD OF DEEP WEB ENTITIES RECOGNITION A CALCULATION METHOD OF DEEP WEB ENTITIES RECOGNITION 1 FENG YONG, DANG XIAO-WAN, 3 XU HONG-YAN School of Informaton, Laonng Unversty, Shenyang Laonng E-mal: 1 fyxuhy@163.com, dangxaowan@163.com, 3 xuhongyan_lndx@163.com

More information

ANALYSIS OF ADAPTIF LOCAL REGION IMPLEMENTATION ON LOCAL THRESHOLDING METHOD

ANALYSIS OF ADAPTIF LOCAL REGION IMPLEMENTATION ON LOCAL THRESHOLDING METHOD Nusantara Journal of Computers and ts Applcatons ANALYSIS F ADAPTIF LCAL REGIN IMPLEMENTATIN N LCAL THRESHLDING METHD I Gust Agung Socrates Ad Guna 1), Hendra Maulana 2), Agus Zanal Arfn 3) and Dn Adn

More information

Determining the Optimal Bandwidth Based on Multi-criterion Fusion

Determining the Optimal Bandwidth Based on Multi-criterion Fusion Proceedngs of 01 4th Internatonal Conference on Machne Learnng and Computng IPCSIT vol. 5 (01) (01) IACSIT Press, Sngapore Determnng the Optmal Bandwdth Based on Mult-crteron Fuson Ha-L Lang 1+, Xan-Mn

More information

Hermite Splines in Lie Groups as Products of Geodesics

Hermite Splines in Lie Groups as Products of Geodesics Hermte Splnes n Le Groups as Products of Geodescs Ethan Eade Updated May 28, 2017 1 Introducton 1.1 Goal Ths document defnes a curve n the Le group G parametrzed by tme and by structural parameters n the

More information

Module Management Tool in Software Development Organizations

Module Management Tool in Software Development Organizations Journal of Computer Scence (5): 8-, 7 ISSN 59-66 7 Scence Publcatons Management Tool n Software Development Organzatons Ahmad A. Al-Rababah and Mohammad A. Al-Rababah Faculty of IT, Al-Ahlyyah Amman Unversty,

More information

Data Modelling and. Multimedia. Databases M. Multimedia. Information Retrieval Part II. Outline

Data Modelling and. Multimedia. Databases M. Multimedia. Information Retrieval Part II. Outline ALMA MATER STUDIORUM - UNIVERSITÀ DI BOLOGNA Data Modellng and Multmeda Databases M Internatonal Second cycle degree programme (LM) n Dgtal Humantes and Dgtal Knowledge (DHDK) Unversty of Bologna Multmeda

More information

BRDPHHC: A Balance RDF Data Partitioning Algorithm based on Hybrid Hierarchical Clustering

BRDPHHC: A Balance RDF Data Partitioning Algorithm based on Hybrid Hierarchical Clustering 015 IEEE 17th Internatonal Conference on Hgh Performance Computng and Communcatons (HPCC), 015 IEEE 7th Internatonal Symposum on Cyberspace Safety and Securty (CSS), and 015 IEEE 1th Internatonal Conf

More information

Feature Reduction and Selection

Feature Reduction and Selection Feature Reducton and Selecton Dr. Shuang LIANG School of Software Engneerng TongJ Unversty Fall, 2012 Today s Topcs Introducton Problems of Dmensonalty Feature Reducton Statstc methods Prncpal Components

More information

OPTIMAL CONFIGURATION FOR NODES IN MIXED CELLULAR AND MOBILE AD HOC NETWORK FOR INET

OPTIMAL CONFIGURATION FOR NODES IN MIXED CELLULAR AND MOBILE AD HOC NETWORK FOR INET OPTIMAL CONFIGURATION FOR NODE IN MIED CELLULAR AND MOBILE AD HOC NETWORK FOR INET Olusola Babalola D.E. Department of Electrcal and Computer Engneerng Morgan tate Unversty Dr. Rchard Dean Faculty Advsor

More information

Web Document Classification Based on Fuzzy Association

Web Document Classification Based on Fuzzy Association Web Document Classfcaton Based on Fuzzy Assocaton Choochart Haruechayasa, Me-Lng Shyu Department of Electrcal and Computer Engneerng Unversty of Mam Coral Gables, FL 33124, USA charuech@mam.edu, shyu@mam.edu

More information

Keyword-based Document Clustering

Keyword-based Document Clustering Keyword-based ocument lusterng Seung-Shk Kang School of omputer Scence Kookmn Unversty & AIrc hungnung-dong Songbuk-gu Seoul 36-72 Korea sskang@kookmn.ac.kr Abstract ocument clusterng s an aggregaton of

More information

BIN XIA et al: AN IMPROVED K-MEANS ALGORITHM BASED ON CLOUD PLATFORM FOR DATA MINING

BIN XIA et al: AN IMPROVED K-MEANS ALGORITHM BASED ON CLOUD PLATFORM FOR DATA MINING An Improved K-means Algorthm based on Cloud Platform for Data Mnng Bn Xa *, Yan Lu 2. School of nformaton and management scence, Henan Agrcultural Unversty, Zhengzhou, Henan 450002, P.R. Chna 2. College

More information

TN348: Openlab Module - Colocalization

TN348: Openlab Module - Colocalization TN348: Openlab Module - Colocalzaton Topc The Colocalzaton module provdes the faclty to vsualze and quantfy colocalzaton between pars of mages. The Colocalzaton wndow contans a prevew of the two mages

More information

Concurrent Apriori Data Mining Algorithms

Concurrent Apriori Data Mining Algorithms Concurrent Apror Data Mnng Algorthms Vassl Halatchev Department of Electrcal Engneerng and Computer Scence York Unversty, Toronto October 8, 2015 Outlne Why t s mportant Introducton to Assocaton Rule Mnng

More information

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour 6.854 Advanced Algorthms Petar Maymounkov Problem Set 11 (November 23, 2005) Wth: Benjamn Rossman, Oren Wemann, and Pouya Kheradpour Problem 1. We reduce vertex cover to MAX-SAT wth weghts, such that the

More information

CS 534: Computer Vision Model Fitting

CS 534: Computer Vision Model Fitting CS 534: Computer Vson Model Fttng Sprng 004 Ahmed Elgammal Dept of Computer Scence CS 534 Model Fttng - 1 Outlnes Model fttng s mportant Least-squares fttng Maxmum lkelhood estmaton MAP estmaton Robust

More information

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices Internatonal Mathematcal Forum, Vol 7, 2012, no 52, 2549-2554 An Applcaton of the Dulmage-Mendelsohn Decomposton to Sparse Null Space Bases of Full Row Rank Matrces Mostafa Khorramzadeh Department of Mathematcal

More information

Lecture 5: Multilayer Perceptrons

Lecture 5: Multilayer Perceptrons Lecture 5: Multlayer Perceptrons Roger Grosse 1 Introducton So far, we ve only talked about lnear models: lnear regresson and lnear bnary classfers. We noted that there are functons that can t be represented

More information

Enhanced Watermarking Technique for Color Images using Visual Cryptography

Enhanced Watermarking Technique for Color Images using Visual Cryptography Informaton Assurance and Securty Letters 1 (2010) 024-028 Enhanced Watermarkng Technque for Color Images usng Vsual Cryptography Enas F. Al rawashdeh 1, Rawan I.Zaghloul 2 1 Balqa Appled Unversty, MIS

More information

3D vector computer graphics

3D vector computer graphics 3D vector computer graphcs Paolo Varagnolo: freelance engneer Padova Aprl 2016 Prvate Practce ----------------------------------- 1. Introducton Vector 3D model representaton n computer graphcs requres

More information

Correlative features for the classification of textural images

Correlative features for the classification of textural images Correlatve features for the classfcaton of textural mages M A Turkova 1 and A V Gadel 1, 1 Samara Natonal Research Unversty, Moskovskoe Shosse 34, Samara, Russa, 443086 Image Processng Systems Insttute

More information

Signature and Lexicon Pruning Techniques

Signature and Lexicon Pruning Techniques Sgnature and Lexcon Prunng Technques Srnvas Palla, Hansheng Le, Venu Govndaraju Centre for Unfed Bometrcs and Sensors Unversty at Buffalo {spalla2, hle, govnd}@cedar.buffalo.edu Abstract Handwrtten word

More information

Video Content Representation using Optimal Extraction of Frames and Scenes

Video Content Representation using Optimal Extraction of Frames and Scenes Vdeo Content Representaton usng Optmal Etracton of rames and Scenes Nkolaos D. Doulam Anastasos D. Doulam Yanns S. Avrths and Stefanos D. ollas Natonal Techncal Unversty of Athens Department of Electrcal

More information

Fuzzy Modeling of the Complexity vs. Accuracy Trade-off in a Sequential Two-Stage Multi-Classifier System

Fuzzy Modeling of the Complexity vs. Accuracy Trade-off in a Sequential Two-Stage Multi-Classifier System Fuzzy Modelng of the Complexty vs. Accuracy Trade-off n a Sequental Two-Stage Mult-Classfer System MARK LAST 1 Department of Informaton Systems Engneerng Ben-Guron Unversty of the Negev Beer-Sheva 84105

More information

Optimal Workload-based Weighted Wavelet Synopses

Optimal Workload-based Weighted Wavelet Synopses Optmal Workload-based Weghted Wavelet Synopses Yoss Matas School of Computer Scence Tel Avv Unversty Tel Avv 69978, Israel matas@tau.ac.l Danel Urel School of Computer Scence Tel Avv Unversty Tel Avv 69978,

More information

y and the total sum of

y and the total sum of Lnear regresson Testng for non-lnearty In analytcal chemstry, lnear regresson s commonly used n the constructon of calbraton functons requred for analytcal technques such as gas chromatography, atomc absorpton

More information