1. Introduction. Abstract

Size: px
Start display at page:

Download "1. Introduction. Abstract"

Transcription

1 Automati Ontology Derivation Uing Clutering for Image Claifiation 1 Latifur Khan and Lei Wang Department of Computer Siene Univerity of Texa at Dalla, TX [lkhan, leiwang]@utdalla.edu Abtrat Tehnology in the field of digital media generate huge amount of non-textual information, audio, video, and image, along with more familiar textual information. The potential for exhange and retrieval of information i vat and daunting. The key problem in ahieving effiient and uer-friendly retrieval in the domain of image i the development of a earh mehanim to guarantee delivery of minimal irrelevant information (high preiion) while inuring that relevant information i not overlooked (high reall). The traditional olution to the problem of image retrieval employ ontentbaed earh tehnique baed on olor, texture or hape feature. The traditional olution work well in performing earhe in whih the uer peifie image ontaining a ample objet, or a ample textural pattern, in whih the objet or pattern i indexed. One an overome thi retrition by indexing image aording to meaning rather than objet that appear in image, although thi will entail a way of onverting objet to meaning. We have olved thi problem of reating a meaning baed index truture through the deign and implementation of a onept-baed model uing domain dependent ontologie. An ontology i a olletion of onept and their interrelationhip whih provide an abtrat view of an appliation domain. With regard to onverting objet to meaning the key iue i to identify appropriate onept that both deribe and identify image. We propoe a new mehanim that an generate ontologie automatially in order to make our approah alable. To ahieve thi we propoe a method for the automati ontrution of ontologie baed on lutering and a vetor pae model. Similarity of image i baed on imilarity of objet that appear in image. For objet imilarity meaure, we onider the ombination of olor and hape imilarity together. 1. Introdution The development of tehnology in the field of digital media generate huge amount of non-textual information, uh a audio, video, and image, a well a more familiar textual information. The potential for the exhange and retrieval of information i vat, and at time daunting. In general, uer an be eaily overwhelmed by the amount of information available via eletroni mean. The need for uer-utomized information eletion i lear. The tranfer of irrelevant information in the form of doument (e.g. text, audio, video) retrieved by an information retrieval ytem and whih are of no ue to the uer wate network bandwidth and frutrate uer. Thi ondition i a reult of inauraie in the repreentation of the doument in the databae, a well a onfuion and impreiion in uer querie, ine uer are frequently unable to expre their need effiiently and aurately. Thee fator ontribute to the lo of information and to the proviion of irrelevant information. Therefore, the key problem to be addreed in information eletion in the domain of image i the development of a earh mehanim whih will guarantee the delivery of a minimum of irrelevant information (high preiion), a well a inuring that relevant information i not overlooked (high reall). 1 Thi tudy wa upported in part by gift from Sun and the National Siene Foundation grant NGS

2 Image onit of variou objet, eah of whih may be ued to effetively laify the image. The untrutured format of image tend to reit tandard ategorization and laifiation tehnique. Traditional ytem ued to tore and proe multimedia image provide no mean of automati laifiation. The ability of thee ytem to retrieve relevant doument baed on earh riteria ould be greatly inreaed if they were able to provide an aurate/emanti deription of an image baed on image ontent. The traditional olution to the problem of image retrieval employ ontent-baed earh tehnique baed on olor, hitogram, texture or hape feature. The traditional olution work well in performing earhe in whih the uer peifie image ontaining a ample objet, or a ample textural pattern. Should a uer ak for an image depiting a baketball game, the reult beome le aurate. Thi i due to the fat that though an image may ontain a baketball, it doe not neearily depit a baketball game. In order to overome the hortoming of traditional tehnique in reponding to image laifiation we have deigned and implemented a onept-baed model uing ontologie. Thi model, whih employ a domain dependent ontology, i preented in thi paper. An ontology i a olletion of onept and their interrelationhip, whih an olletively provide an abtrat view of an appliation domain. In our ytem we would like to addre two ditint quetion: the extration of the emanti onept from the image and the ontrution of an ontology. With regard to the firt problem, the extration of emanti onept, the key iue i to identify appropriate onept that deribe and identify image. We would like to make ure that irrelevant onept will not be aoiated and mathed, and that relevant onept will not be diarded. In other word, it i important to inure that high preiion and high reall will be preerved during onept eletion. To the bet of our knowledge there are no attempt to onnet image and onept through the ue of ontologie in any traditional image retrieval ytem. We propoe an automati mehanim for the eletion of thee onept [16]. With regard to the eond problem, we propoe a new method for the automati ontrution of ontologie baed on lutering and vetor pae model. It i important to note that imilarity of image i baed on imilarity of objet that appear in image. In addition, objet imilarity take into aount not only olor or hape, but both. Our method ontrut ontologie automatially in bottom up fahion. For thi, we firt ontrut a hierarhy uing ome lutering algorithm. Reall that if doument are imilar to eah other in ontent they will be aoiated with the ame onept in ontology. After objet detetion, for eah objet, we extrat olor and hape information and expre them uing vetor. Then onidering both olor and hape fator, we an alulate the imilarity between objet applying vetor pae model. In ontologie eah onept i deribed by a et of feature (objet). Thu, before image lutering, we luter objet aording to imilaritie between objet and aign a weight for eah objet luter. Next, we ontrut a vetor for eah image baed on weight of objet luter and alulate imilaritie between image uing vetor pae model. Finally, baed on image imilaritie, we luter image and build ontology hierarhy uing hierarhy agglomerative lutering algorithm. Setion of thi paper diue work related to ontent-baed image retrieval and ontologie for ue in image retrieval, a well a the urrent ytem ued for image proeing. Setion 3 deribe ontologie, and how they peify interrelationhip among onept that help draw meaningful onluion about image. Setion 4 deribe outline of our approah. Setion 5 preent elaborately our approah to luter image and build ontology hierarhy. Setion 6 preent preliminary reult of our approah. Setion 7 preent our onluion and poible area of future work.

3 . Related Work Several ytem exit today that attempt to laify image baed on their ontent. Sueful laifiation of an image and it ontent relate diretly to how well relevant image may be retrieved when a earh i preformed. Mot image toring ytem uh a QBIC and ViualSEEK limit laifiation mehanim to deribing an image baed on metadata uh a olor hitogram, texture, or hape feature [, 8]. Thee ytem have high ue in performing earhe in whih the uer peifie image ontaining a ample objet, or a ample texture pattern. Should a uer ak for an image depiting a baketball game, the reult beome le aurate. Thi i due to the fat that though an image may ontain a baketball, it doe not depit a baketball game. Sytem that only ontain metadata regarding only olor and hape feature ontained in an image annot provide an aurate laifiation of the entire image. Other ytem attempt to provide image with a more preie deription by analyzing other element urrounding the image, uh a aption [9, 10], or HTML tag on web page [1]. Thee ytem ue thi information to help laify the image and give it a meaningful deription. Thi approah, tied together with metadata on image uh a texture, and olor ampling ha the potential to yield high preiion reult in image laifiation. Examining the textual deription aoiated with an image provide additional information that may be ued to help better laify the image. Unfortunately, thi approah doe not take into aount the onnetion among individual objet preent in a ample image. Suh onnetion provide ueful information in the form of relationhip among objet preent in the image, whih ould be ued to help laify the image ontent. For the ontrution of ontologie, only a few automati method are propoed [4, 6, 7]. Elliman et al. [7] propoe a method for ontruting ontologie to repreent a et of web page on a peified ite. Self organizing map i ued to ontrut hierarhy. In our ae we modify elf organizing tree and label node in the hierarhy. Bodner et al. [4] propoe a method to ontrut hierarhy baed on tatitial method (frequeny of word). Hoothe et al. [6] propoe variou lutering tehnique to view text doument with the help of ontologie. Note that a et of hierarhie will be ontruted for multiple view only; not for ontology ontrution purpoe. Furthermore, all thee ontology ontrution are done in text domain; however, we addre thi problem in the image domain. In our ytem, ontology erve a a taxonomy where imilar image are grouped together [3, 16]. Thi imilarity i not baed only on olor or hape but both along with finer grain (i.e., individual objet) rather than oarer grain (i.e., entire image). For example, a football game image may ontain green field, goalpot, and football objet. An image ontaining only a football would be milaified a a football game baed on olor imilarity analyi. On the other hand, hape imilarity may alo milaify image. Baed on only hape imilarity we may identify a baketball a a football. Therefore, neither olor nor hape baed imilarity i adequate to laify image. We need to ombine thee two imilaritie together to undertand emanti meaning of image. Therefore, to laify image effetively, we need a knowledgebae where olor and hape feature of eah ategory will be maintained. In our ae ontology erve a a knowledgebae; it ontain a et of onept where eah onept orrepond a ategory. And eah ategory ontain a et of image by haring a et of imilar objet. Here image imilarity i determined by objet imilarity baed on the ombination of olor and hape. One may argue that our ontology generation i baed on lutering apet of the problem. However, thi lutering group imilar image baed on emanti meaning. Thu, onept-baed lutering tehnique ha been employed.

4 3. Ontologie An ontology i a peifiation of an abtrat, implified view of the world that we wih to repreent for ome purpoe. Therefore, an ontology define a et of repreentational term that we all onept. Inter-relationhip among thee onept deribe a target world. An ontology an be ontruted in two way, domain dependent and generi. CYC, WordNet, and Senu are example of generi ontologie. For our purpoe, we hooe a domain-dependent ontology. A domain-dependent ontology provide onept in a fine grain, while generi ontologie provide onept in oarer grain. The finegrained onept allow u to determine peifi relationhip among feature in image that may be ued to effetively laify thoe image. Figure 1 illutrate an example ontology for the port domain [5]. Thi ontology may be obtained from generi port terminology and domain expert. The ontology i deribed by a direted ayli graph (DAG). Here, eah node in the DAG repreent a onept. In general, eah onept in the ontology ontain a label name and feature vetor. A feature vetor i imply a et of feature and their weight. Eah feature may repreent an objet of an image, uh a a baketball, a goalpot or a baeball. Note alo that thi label name onneted to the feature i unique in the ontology. Furthermore, thi label name i ued to erve a an aoiation of onept to image. The onept of football may be further expanded to objet preent in a football game (i.e. the feature of the onept). For intane, a green field, goalpot, and football player would indiate the image i a football game. Should only one or two of the feature ommon to a football game (a peified in the ontology) be preent, a le peifi laifiation of the image would be given. In other word, a more generi onept will be aigned to the image. An image ontaining only a football would be laified a an image ontaining a football, not a a football game. Furthermore, the weight of eah feature of a onept may not be equal. In other word, for a partiular onept ome feature may erve a more diriminating a ompared to ome other; it will be aigned higher weight. For example, in the onept of a game of football the weight of goalpot feature i higher than the weight of the feature, green field. In ontologie, onept are interonneted by mean of inter-relationhip. If there i a inter-relationhip R, between onept C i and C j, then there i alo a inter-relationhip R between onept C j and C i. In Figure 1, inter-relationhip are repreented by labeled ar/link. Three kind of inter-relationhip are ued to reate our ontology: IS-A, Intane-of, and Part-of. Thee orrepond to key abtration primitive in objet-baed and emanti data model [1]. Figure 1: A Portion of an ontology for the Sport Domain

5 4. Propoed Sytem Our ytem irumvent the low preiion laifiation tehnique of other ytem by examining the atual objet within an image and uing them to diover relationhip that reveal information ueful in laifying the entire image. The onept behind thee relationhip are held in our knowledge bae of domain-dependant ontologie a deribed in etion 3. We now outline the tep taken to uefully proe and laify an input image preented to our ytem. To onvert objet to meaning or automati ontology ontrution, we need to identify all objet boundarie aurately (box 1 in Figure 1) that appear in image [3]. In thi earlier work [3] we propoe an automati alable objet boundary detetion algorithm baed on edge detetion and region growing tehnique. Our algorithm work in three tage. Firt, we detet all edge pixel in image and divide pixel into two et, edge pixel and region pixel et. Seond, we grow a region from the region pixel et urrounded by edge taken from the edge pixel et. Finally, we may merge adjaent region uing an adjaeny graph to avoid over egmentation of region and to detet boundary of objet aurately. After detetion of objet boundarie, we would like to build ontologie automatially (box in Figure ). Our method ontrut ontologie automatially in bottom up fahion. For thi, we firt ontrut a hierarhy uing ome lutering algorithm. Reall that if doument are imilar to eah other in ontent they will be aoiated with the ame onept in ontology. After objet detetion, for eah objet, we extrat olor and hape information and expre them uing vetor. Then onidering both olor and hape fator, we an alulate the imilarity between objet applying vetor pae model. In ontologie eah onept i deribed by a et of feature (objet). Thu, before image lutering, we luter objet aording to imilaritie between objet and aign a weight for eah objet luter. Next, we ontrut a vetor for eah image baed on weight of objet luter and alulate imilaritie between image uing vetor pae model. Finally, baed on image imilaritie, we luter image and build ontology hierarhy uing hierarhy agglomerative lutering algorithm. Here a et of image will be ued for ontrution of hierarhy. Thu, after ontrution of hierarhy, all emantially imilar image baed on objet imilarity will be grouped together. Furthermore, to laify a query image, firt we egment image into objet baed on our objet boundary detetion algorithm. Next, we determine imilarity between objet that appear in thi query image and objet that appear in onept entroid image uing vetor pae model (box 3 in Figure ), and hooe the mot imilar one [16]. After the objet have been identified, their identifiation are fed into a onept eletion module (box 4 in Figure ). The ontologie ue thi information to provide a meaningful deription of the image by eleting onept baed on image ontent (i.e., individual objet within the image). In our earlier work [15] onept eletion mehanim inlude a novel, alable diambiguation algorithm uing a domain peifi ontology. Thi algorithm will prune irrelevant onept while allowing relevant onept to beome aoiated with image. For example, it i poible an image may be laified a both an NBA baketball game and a ollege baketball game at the ame time. However, we employ the heuriti-baed pruning tehnique to narrow down the eletion of onept. When the pruning algorithm omplete the eleted onept will be orted baed on their ranking in deending order. The onept label will then tell whih ategory the image belong to.

6 Training Image (TI) Objet Detetion (I) (I) Objet in TI Ontology Contrution Objet aoiated in a onept Similarity meaure of objet uing Vetor Spae Model Objet of QI appeared in onept Conept Seletion Query Image (QI) Objet in QI Figure. Flow of Our Sytem 5. Ontology Derivation Our goal i to ontrut ontologie automatially. For thi, we would like to build a hierarhy from a et of image in a bottom up fahion. Note that we do not want to luter image baed on the imilarity of olor. Hene, imilar image will be lutered if they hare imilar objet. Furthermore, ome objet may arry more weight a ompared to other objet that appear in image, imilar to the way keyword behave in doument. For thi, we need to aign weight to objet that appear in image. To determine the weight of objet we firt luter objet baed on imilarity of olor and hape. Next, we determine weight of objet that appear in image baed on term frequeny and invere doument frequeny, imilar to IR. Eah image will then be repreented by a vetor, where a vetor ontain a et of weight that orrepond to the importane of the objet that appear in the image. Uing thi vetor we will ontrut an image hierarhy uing agglomerative lutering. We will diu eah tage elaborately in the following etion: 5.1 Objet Clutering By applying egmentation tehnique, we egment image into objet. Let u aume we have N image in a databae. After egmentation, we detet M number of objet in total. Then we luter thee M objet into a et of group (ay t) C 1, C,. C t aording to imilaritie between objet. Here, objet imilarity will be baed on the ombination of olor imilarity and hape imilarity of individual objet. Thi i beaue if viual feature of objet uh a olor and hape are imilar it i very poible that thee objet have imilar emanti meaning. Thu, we firt introdue our olor imilarity meaure, next our hape imilarity meaure, and then the ombination of both Color Similarity Meaure To ompute olor imilarity, we firt ontrut a vetor for an objet oniting of a et of value of hitogram bin. For eah hitogram bin (olor ode), we determine how many pixel of thi partiular objet appear in the hitogram bin. Thu, for objet i a vetor Vi (v 1,i, v,i,.. v p,i, v k,i ) will be ontruted to expre the olor hitogram. In the vetor eah element repreent the perentage of pixel whoe hue value loate in peifi interval. For example, v p,i i the perentage of pixel whoe hue value are between * π * p / k and * π * (p + 1) / k in objet i, beaue the range of hue value in HSI olor pae i from 0 to * π. Furthermore, we only onider hue omponent for imilarity

7 meaure whih i adequate. Now, we an determine the degree of olor imilarity (im (i, ) between objet i and objet j baed on oine produt. Thu, ρ ρ k i j v p = p i v 1, p, j im ( i, = ρ ρ = k k i j v p = p i v 1, p = 1 p, j (1) The value of k affet the auray of olor imilarity. Along with inreaing of k, auray will be alo inreaed. However, it will be omputationally expenive Shape Similarity Meaure The hape imilarity i little bit more ompliated. To upport imilarity querie, Lu et al. [14] introdue fixed reolution (FR) repreentation method. We adopt thi idea, but make ome hange during implementation. Firt, we need to find the major axi that i the longet line joining two point on the boundary. For normalization purpoe we rotate an angle θ around ma entroid of objet to make the major axi to be parallel to x-axi and keep the entroid above the major axi. The reaon for normalization i to make it invariant to rotation [13]. The oordinate of the entroid are a follow: σ(x,y) i urfae denity funtion. () After normalization, we divide the objet into q * q grid, whih i jut big enough to over the entire objet, and overlaid on the objet where q i an integer number. The ize of eah ell i ame. Then we define a hape vetor for objet i, Ui (u 1,i, u,i,.. u p,i, u q,i ) of q ize. Eah element in the vetor tand for the perentage of pixel in the orreponding ell. The higher the q value, the higher the auray. Of oure, it will then be omputationally expenive. Finally, we determine hape imilarity (im (i, ) between two objet i and j uing oine imilarity. Thu, im ρ ρ q i j u p = 1 p, i u p, j ( i, = ρ ρ = i q q j u p = 1 p, i u p = 1 (3) Combined Similarity Now, we would like to determine imilarity between two objet baed on olor imilarity and hape imilarity. Uing Equation and 3, imilarity (im (i, ) between objet i and objet j i a follow: im( i, = im weight + weight ( i, weight = 1 + im ( i, weight (4) Weight i the weight of olor and weight i the weight of hape. When it i poible that one type of imilarity may be more important a ompared to another we need to ue weight. In the urrent ae, we aume that both weight and weight are equally important (=0.5). To ontrut objet luter we ue a threhold T obj. If imilarity between two objet i greater than T obj, then the two objet an be in the ame group. In other word, imilarity between eah pair of objet in the ame group mut be greater than T obj. Furthermore, it i poible for an objet to appear in more than one group. It i important to note that a group oniting of a et of objet orrepond to a keyword a oppoed to eah individual objet orreponding to a keyword. Thi i beaue in IR math i well defined; in multimedia, math i ill defined (imilarity). p, j

8 5. Vetor Model for Image Image lutering i baed on image imilarity. To alulate image imilarity, we ontrut an image, l vetor W l (w 1,l, w,l w i,l, w t,l ) rather than meauring imilarity baed on olor. W i,l i the weight of objet luter C i in image l. In thi vetor we keep the weight of eah group. Thu, the ize of vetor W i ame a the total number of objet luter (=t). It i poible that the weight of a group may be zero. Thi i beaue no objet of an image may be a partiipant of that group during lutering. To determine an image vetor, we adopt the idea from the area of information retrieval [39]. Here, image orrepond to doument, and objet luter orrepond to term (keyword). Let N be the total number of image and n i be the number of image in whih objet of luter C i appear. We define the normalized term frequeny f i,j : f i, l freqi, l max freq = (5) h h, l The freq i,l i the number of time luter C i appear in the image l. Similarly, for max freq h,l we determine ourrene of a luter that appear in image, l; for thi luter we get maximum ourrene among all luter. We alo define an invere doument frequeny idf i for C i : idf i N log n = (6) i Conidering the above two fator, we have the weight of luter: w f idf i, l i, l i = i, l f N log n = (7) i After omputing image vetor, we get imilarity between any two image uing oine imilarity: ρ ρ t i j w h = h i w 1, h, j im img ( i, = ρ ρ = t t i j (8) w h = h i w 1, h = 1 h, j 5.3 Hierarhial Clutering of Image Uing the above method, we an alulate image imilarity between eah pair of image. Then we apply a hierarhial agglomerative lutering algorithm (HAC) to ontrut hierarhy. The HAC algorithm i a ommonly employed laial hierarhal lutering algorithm. The reult of HAC i a dendrogram repreenting the neted grouping of image. The general HAC algorithm i a follow: 1) Put eah image into a ingleton luter, ompute a lit of inter luter ditane for all ingleton luter, then ort the lit in aending order. ) Find the pair of luter with the mot imilar, merge them into one luter and alulate the imilarity between the new luter and the remaining luter. 3) While there i more than one luter remaining, go to tep, otherwie top. Baed on the alulation of imilarity between the non-ingleton luter a variety of hierarhial agglomerative tehnique have been propoed. Single-link, omplete-link and groupaverage-link lutering are ommonly ued. In the ingle-link luter the imilarity between two luter i the maximum imilarity of all pair of doument whih are in different luter. In the omplete-link luter, the imilarity between two luter i the minimum imilarity of all pair of doument whih are in different luter. In Group-Average-link lutering the imilarity between two luter i the mean imilarity of all pair of ingleton whih are in different luter.

9 In the hierarhy, a node will be repreented by a repreentative image whih i mot imilar to the node vetor. It i important to note that variou type of inter-relationhip between node are blurred in our ontologie; ertain type of interonnetion are ignored. Thi i beaue our prime onern i to failitate information eletion rather than to dedut new knowledge. 6. Experimental Preliminary Reult The purpoe of the experiment i to tet the auray of the lutering. We have ued a et of image belonging to 6 different ategorie, uh a baketball game, baeball game, bat, football game, goggle and playground. We have then ontruted hierarhy baed on the theory diued in Setion 5. We have hoen Group-Average-Link lutering in HAC. We have et k=1 (ee Equation 1), and T obj =0.7. To meaure the quality of a luter, we ue preiion, reall, and E meaure [33]. Reall i the ratio of relevant image to total image for a given ategory. Preiion i the ratio of relevant image to image that appear in a luter for a given ategory. E meaure i defined a follow: E( p, r) = 1 (9) 1 p + 1 r Where p and r are the Preiion and Reall of a luter. Note that E (p, r) i imply one minu harmoni mean of the preiion and reall; E (p, r) range from 0 to 1 where E (p, r) =0 orrepond to perfet preiion and reall, and E (p, r) orrepond to zero preiion and reall. Thu, the maller the E meaure value the better the quality of a luter. In Figure 3 X axi repreent different ategorie and the Y axi repreent p, r and E. We have oberved that preiion and reall are higher for imple image (e.g., play ground) a ompared to preiion and reall of omplex image (e.g., baeball game, football game). Thu, E value i lower for imple image (e.g., play ground) a ompared to E value of omplex image (e.g., baeball game, football game). preiion p,r and E reall ommon E meaure baketball football baeball bat goggle playground Categorie 7. Conluion and Future Work Figure 3. Cluter Quality for Different Categorie In thi paper we have propoed a potentially powerful and novel approah for the automati ontrution of ontology. The rux of our innovation i the development of a hierarhy baed on objet imilarity uing a vetor pae model. Furthermore, to determine objet imilarity we have ombined both olor imilarity and hape imilarity. To illutrate the effetivene of our algorithm in automati

10 image laifiation, we implement a very bai ytem aimed at the laifiation of image in the port domain. For developing a hierarhy, we have ued an agglomerative lutering algorithm that ontrut hierarhie from bottom to up. We would like to extend thi work in the following diretion. Firt, we would like to do more experimentation with the lutering tehnique. Next, we would like to addre thi kind of ontology ontrution in variou domain. Referene [1] G. Alan and D. MLeod, Semanti Heterogeneity Reolution in Federated Databae by Metadata Implantation and Stepwie Evolution, The International Journal on Very Large Databae, Vol. 18, No., Otober [] R. Barber, W. Equitz, C. Falouto, M. Fikner, W. Niblak, D. Petkovi, and P. Yanker, Query by Content for Large On-Line Image Colletion, IEEE Journal, [3] Lei Wang, Latifur Khan, and Caey Breen, Objet Boundary Detetion for Ontology-baed Image Claifiation, Third International Workhop on Multimedia Data Mining, Edmonton, Alberta, Canada, July 00. [4] R. Bodner and F. Song, Knowledge-baed Approahe to Query Expanion in Information Retrieval, in Pro. of Advane in Artifiial Intelligene, pp , New York, Springer. [5] ESPN CLASSIC, [6] Dave Elliman, J. Rafael G. Pulido. Automati Derivation of On-line Doument Ontology. International Workhop on Mehanim for Enterprie Integration: From Objet to Ontology (MERIT 001) 15th European Conferene on Objet Oriented Programming, Budapet, Hungary, Jun 001. [7] A. Hotho, A. Mädhe, A., S. Staab, Ontology-baed Text Clutering, Workhop Text Learning: Beyond Superviion, 001. [8] A. Pentland, R.W. Piard, S. Slaroff, Photobook: Tool for Content-Baed Manipulation of Image Databae, in Pro. of Storage and Retrieval for Image and Video Databae II, Volume 185, pp , Bellingham, WA, [9] N. Row, and B. Frew, Automati Claifiation of Objet in Captioned Depitive Photograph for Retrieval, Intelligent Multimedia Information Retrieval, Chapter 7, M. Maybury, AAAI Pre, [10] A. F. Smeaton and A. Quigley, Experiment on Uing Semanti Ditane between Word in Image Caption Retrieval, in Pro. of The Nineteenth Annual International ACM SIGIR Conferene on Reearh and Development in Information Retrieval, [11] C.J. van Rijbergen. Information Retrieval. Butterworth, London, [1] C. Frankel, M.J. Swain and V. Athito, WebSeer: An Image Searh Engine for the World Wide Web, Univerity of Chiago Tehnial Report TR-96-14, July 31, [13] Chakrabarti, K., Ortega-Binderberger, M., Porkaew, K & Mehrotra, S. (000) Similar hape retrieval in MARS. Proeeding of IEEE International Conferene on Multimedia and Expo. [14] G. Lu and A. Sajjanhar, Region-baed hape repreentation and imilarity meaure uitable for ontent-baed image retrieval. Springer Verlag Multimedia Sytem, [15] Riardo Baeza-Yate, Berthier Ribeiro-Neto, Modern Information Retrieval, ISBN X, [16] C. Breen, L. Khan, A. Ponnuamy, and L. Wang, Ontology-baed Image Claifiation Uing Neural Network, Pro. of SPIE Internet Multimedia Management Sytem III, pp , Boton, MA, July, 00.

KINEMATIC ANALYSIS OF VARIOUS ROBOT CONFIGURATIONS

KINEMATIC ANALYSIS OF VARIOUS ROBOT CONFIGURATIONS International Reearh Journal of Engineering and Tehnology (IRJET) e-in: 39-6 Volume: 4 Iue: May -7 www.irjet.net p-in: 39-7 KINEMATI ANALYI OF VARIOU ROBOT ONFIGURATION Game R. U., Davkhare A. A., Pakhale..

More information

Description of Traffic in ATM Networks by the First Erlang Formula

Description of Traffic in ATM Networks by the First Erlang Formula 5th International Conferene on Information Tehnology and Appliation (ICITA 8) Deription of Traffi in ATM Network by the Firt Erlang Formula Erik Chromý, Matej Kavaký and Ivan Baroňák Abtrat In the paper

More information

Visual Targeted Advertisement System Based on User Profiling and Content Consumption for Mobile Broadcasting Television

Visual Targeted Advertisement System Based on User Profiling and Content Consumption for Mobile Broadcasting Television Viual Targeted Advertiement Sytem Baed on Uer Profiling and ontent onumption for Mobile Broadating Televiion Silvia Uribe Federio Alvarez Joé Manuel Menéndez Guillermo inero Abtrat ontent peronaliation

More information

Using Bayesian Networks for Cleansing Trauma Data

Using Bayesian Networks for Cleansing Trauma Data Uing Bayeian Network for Cleaning Trauma Data Prahant J. Dohi pdohi@.ui.edu Dept. of Computer Siene Univ of Illinoi, Chiago, IL 60607 Lloyd G. Greenwald lgreenwa@.drexel.edu Dept. of Computer Siene Drexel

More information

Incorporating Speculative Execution into Scheduling of Control-flow Intensive Behavioral Descriptions

Incorporating Speculative Execution into Scheduling of Control-flow Intensive Behavioral Descriptions Inorporating Speulative Exeution into Sheduling of Control-flow Intenive Behavioral Deription Ganeh Lakhminarayana, Anand Raghunathan, and Niraj K. Jha Dept. of Eletrial Engineering C&C Reearh Laboratorie

More information

ISSN (Online), Volume 1, Special Issue 2(ICITET 15), March 2015 International Journal of Innovative Trends and Emerging Technologies

ISSN (Online), Volume 1, Special Issue 2(ICITET 15), March 2015 International Journal of Innovative Trends and Emerging Technologies International Journal of Innovative Trend and Emerging Tehnologie ROBUST SCAN TECHNIQUE FOR SECURED AES AGAINST DIFFERENTIAL CRYPTANALYSIS BASED SIDE CHANNEL ATTACK A.TAMILARASAN 1, MR.A.ANBARASAN 2 1

More information

Combined Radix-10 and Radix-16 Division Unit

Combined Radix-10 and Radix-16 Division Unit Combined adix- and adix-6 Diviion Unit Tomá ang and Alberto Nannarelli Dept. of Eletrial Engineering and Computer Siene, Univerity of California, Irvine, USA Dept. of Informati & Math. Modelling, Tehnial

More information

Datum Transformations of NAV420 Reference Frames

Datum Transformations of NAV420 Reference Frames NA4CA Appliation Note Datum ranformation of NA4 Referene Frame Giri Baleri, Sr. Appliation Engineer Crobow ehnology, In. http://www.xbow.om hi appliation note explain how to onvert variou referene frame

More information

Relayer Selection Strategies in Cellular Networks with Peer-to-Peer Relaying

Relayer Selection Strategies in Cellular Networks with Peer-to-Peer Relaying Relayer Seletion Strategie in Cellular Network with Peer-to-Peer Relaying V. Sreng, H. Yanikomeroglu, and D. D. Faloner Broadband Communiation and Wirele Sytem (BCWS) Centre Dept. of Sytem and Computer

More information

Q1:Choose the correct answer:

Q1:Choose the correct answer: Q:Chooe the orret anwer:. Purpoe of an OS i a. Create abtration b. Multiple proee ompete for ue of proeor. Coordination. Sheduler deide a. whih proee get to ue the proeor b. when proee get to ue the proeor.

More information

Inverse Kinematics 1 1/29/2018

Inverse Kinematics 1 1/29/2018 Invere Kinemati 1 Invere Kinemati 2 given the poe of the end effetor, find the joint variable that produe the end effetor poe for a -joint robot, given find 1 o R T 3 2 1,,,,, q q q q q q RPP + Spherial

More information

COURSEWORK 1 FOR INF2B: FINDING THE DISTANCE OF CLOSEST PAIRS OF POINTS ISSUED: 9FEBRUARY 2017

COURSEWORK 1 FOR INF2B: FINDING THE DISTANCE OF CLOSEST PAIRS OF POINTS ISSUED: 9FEBRUARY 2017 COURSEWORK 1 FOR INF2B: FINDING THE DISTANCE OF CLOSEST PAIRS OF POINTS ISSUED: 9FEBRUARY 2017 Submiion Deadline: The ourework onit of two part (of a different nature) relating to one problem. A hown below

More information

Automatic design of robust PID controllers based on QFT specifications

Automatic design of robust PID controllers based on QFT specifications IFAC Conferene on Advane in PID Control PID'1 Breia (Italy), Marh 8-3, 1 Automati deign of robut PID ontroller baed on QFT peifiation R. Comaòliva*, T. Eobet* J. Quevedo* * Advaned Control Sytem (SAC),

More information

Parametric Micro-level Performance Models for Parallel Computing

Parametric Micro-level Performance Models for Parallel Computing Computer Siene Tehnial Report Computer Siene 12-5-1994 Parametri Miro-level Performane Model for Parallel Computing Youngtae Kim Iowa State Univerity Mark Fienup Iowa State Univerity Jeffrey S. Clary Iowa

More information

Macrohomogenous Li-Ion-Battery Modeling - Strengths and Limitations

Macrohomogenous Li-Ion-Battery Modeling - Strengths and Limitations Marohomogenou Li-Ion-Battery Modeling - Strength and Limitation Marku Lindner Chritian Wieer Adam Opel AG Sope Purpoe of the reearh: undertand and quantify impat of implifiation in marohomogeneou model

More information

An Evolutionary Multiple Heuristic with Genetic Local Search for Solving TSP

An Evolutionary Multiple Heuristic with Genetic Local Search for Solving TSP An Evolutionary Multiple Heuriti with Geneti Loal Searh for Solving TSP Peng Gang Ihiro Iimura 2 and Shigeru Nakayama 3 Department of Information and Computer Siene Faulty of Engineering Kagohima Univerity

More information

Universität Augsburg. Institut für Informatik. Approximating Optimal Visual Sensor Placement. E. Hörster, R. Lienhart.

Universität Augsburg. Institut für Informatik. Approximating Optimal Visual Sensor Placement. E. Hörster, R. Lienhart. Univerität Augburg à ÊÇÅÍÆ ËÀǼ Approximating Optimal Viual Senor Placement E. Hörter, R. Lienhart Report 2006-01 Januar 2006 Intitut für Informatik D-86135 Augburg Copyright c E. Hörter, R. Lienhart Intitut

More information

OSI Model. SS7 Protocol Model. Application TCAP. Presentation Session Transport. ISDN-UP Null SCCP. Network. MTP Level 3 MTP Level 2 MTP Level 1

OSI Model. SS7 Protocol Model. Application TCAP. Presentation Session Transport. ISDN-UP Null SCCP. Network. MTP Level 3 MTP Level 2 MTP Level 1 Direte Event Simulation of CCS7 DAP Benjamin, AE Krzeinki and S Staven Department of Computer Siene Univerity of Stellenboh 7600 Stellenboh, South Afria fbenj,aek,taveng@.un.a.za ABSTRACT: Complex imulation

More information

MAT 155: Describing, Exploring, and Comparing Data Page 1 of NotesCh2-3.doc

MAT 155: Describing, Exploring, and Comparing Data Page 1 of NotesCh2-3.doc MAT 155: Decribing, Exploring, and Comparing Data Page 1 of 8 001-oteCh-3.doc ote for Chapter Summarizing and Graphing Data Chapter 3 Decribing, Exploring, and Comparing Data Frequency Ditribution, Graphic

More information

Volume 3, Issue 9, September 2013 International Journal of Advanced Research in Computer Science and Software Engineering

Volume 3, Issue 9, September 2013 International Journal of Advanced Research in Computer Science and Software Engineering Volume 3, Issue 9, September 2013 ISSN: 2277 128X International Journal of Advaned Researh in Computer Siene and Software Engineering Researh Paper Available online at: www.ijarsse.om A New-Fangled Algorithm

More information

Pruning Game Tree by Rollouts

Pruning Game Tree by Rollouts Pruning Game Tree by Rollout Bojun Huang Mirooft Reearh bojhuang@mirooft.om Abtrat In thi paper we how that the α-β algorithm and it ueor MT-SSS*, a two lai minimax earh algorithm, an be implemented a

More information

A Novel Method for Removing Image Staircase Artifacts

A Novel Method for Removing Image Staircase Artifacts Vol.139 (SIP 016, pp.56-63 http://dx.doi.org/10.1457/atl.016.139.54 A Novel Method for Removing Image Stairae Artifat Zhong Chen 1,, Zhiwei Hou 1, Yuegang Xing, Xiaobing Chen 1 1 Jiangu Key Laboratory

More information

Pipelined Multipliers for Reconfigurable Hardware

Pipelined Multipliers for Reconfigurable Hardware Pipelined Multipliers for Reonfigurable Hardware Mithell J. Myjak and José G. Delgado-Frias Shool of Eletrial Engineering and Computer Siene, Washington State University Pullman, WA 99164-2752 USA {mmyjak,

More information

Shortest Paths in Directed Graphs

Shortest Paths in Directed Graphs Shortet Path in Direted Graph Jonathan Turner January, 0 Thi note i adapted from Data Struture and Network Algorithm y Tarjan. Let G = (V, E) e a direted graph and let length e a real-valued funtion on

More information

ABSTRACT 1. INTRODUCTION

ABSTRACT 1. INTRODUCTION Object Boundary Detection for Ontology-based Image Classification * Lei Wang, Latifur Khan, and Casey Breen Department of Computer Science University of Texas at Dallas, TX 75083-0688 Email: [leiwang,

More information

View-Based Tree-Language Rewritings

View-Based Tree-Language Rewritings View-Baed Tree-Language Rewriting Lak Lakhmanan, Alex Thomo Univerity of Britih Columbia, Canada Univerity of Vitoria, Canada Importane of tree XML Semi-trutured textual format are very popular.

More information

Laboratory Exercise 6

Laboratory Exercise 6 Laboratory Exercie 6 Adder, Subtractor, and Multiplier The purpoe of thi exercie i to examine arithmetic circuit that add, ubtract, and multiply number. Each type of circuit will be implemented in two

More information

Interconnection Styles

Interconnection Styles Interonnetion tyles oftware Design Following the Export (erver) tyle 2 M1 M4 M5 4 M3 M6 1 3 oftware Design Following the Export (Client) tyle e 2 e M1 M4 M5 4 M3 M6 1 e 3 oftware Design Following the Export

More information

Calculations for multiple mixers are based on a formalism that uses sideband information and LO frequencies: ( ) sb

Calculations for multiple mixers are based on a formalism that uses sideband information and LO frequencies: ( ) sb Setting frequeny parameter in the WASP databae A. Harri 24 Aug 2003 Calulation for multiple mixer are baed on a formalim that ue ideband information and LO frequenie: b b := ign f ig f LO f IF := f ig

More information

Contents Contents...I List of Tables...VIII List of Figures...IX 1. Introduction Information Retrieval... 8

Contents Contents...I List of Tables...VIII List of Figures...IX 1. Introduction Information Retrieval... 8 Contents Contents...I List of Tables...VIII List of Figures...IX 1. Introdution... 1 1.1. Internet Information...2 1.2. Internet Information Retrieval...3 1.2.1. Doument Indexing...4 1.2.2. Doument Retrieval...4

More information

Performance of Histogram-Based Skin Colour Segmentation for Arms Detection in Human Motion Analysis Application

Performance of Histogram-Based Skin Colour Segmentation for Arms Detection in Human Motion Analysis Application World Aademy of Siene, Engineering and Tehnology 8 009 Performane of Histogram-Based Skin Colour Segmentation for Arms Detetion in Human Motion Analysis Appliation Rosalyn R. Porle, Ali Chekima, Farrah

More information

NONLINEAR BACK PROJECTION FOR TOMOGRAPHIC IMAGE RECONSTRUCTION. Ken Sauer and Charles A. Bouman

NONLINEAR BACK PROJECTION FOR TOMOGRAPHIC IMAGE RECONSTRUCTION. Ken Sauer and Charles A. Bouman NONLINEAR BACK PROJECTION FOR TOMOGRAPHIC IMAGE RECONSTRUCTION Ken Sauer and Charles A. Bouman Department of Eletrial Engineering, University of Notre Dame Notre Dame, IN 46556, (219) 631-6999 Shool of

More information

Transition Detection Using Hilbert Transform and Texture Features

Transition Detection Using Hilbert Transform and Texture Features Amerian Journal of Signal Proessing 1, (): 35-4 DOI: 1.593/.asp.1.6 Transition Detetion Using Hilbert Transform and Texture Features G. G. Lashmi Priya *, S. Domni Department of Computer Appliations, National

More information

Topics. Lecture 37: Global Optimization. Issues. A Simple Example: Copy Propagation X := 3 B > 0 Y := 0 X := 4 Y := Z + W A := 2 * 3X

Topics. Lecture 37: Global Optimization. Issues. A Simple Example: Copy Propagation X := 3 B > 0 Y := 0 X := 4 Y := Z + W A := 2 * 3X Lecture 37: Global Optimization [Adapted from note by R. Bodik and G. Necula] Topic Global optimization refer to program optimization that encompa multiple baic block in a function. (I have ued the term

More information

DAROS: Distributed User-Server Assignment And Replication For Online Social Networking Applications

DAROS: Distributed User-Server Assignment And Replication For Online Social Networking Applications DAROS: Ditributed Uer-Server Aignment And Replication For Online Social Networking Application Thuan Duong-Ba School of EECS Oregon State Univerity Corvalli, OR 97330, USA Email: duongba@eec.oregontate.edu

More information

SPH3UW Unit 7.1 The Ray Model of Light Page 2 of 5. The accepted value for the speed of light inside a vacuum is c m which we usually

SPH3UW Unit 7.1 The Ray Model of Light Page 2 of 5. The accepted value for the speed of light inside a vacuum is c m which we usually SPH3UW Unit 7. The Ray Model of Light Page of 5 Note Phyi Tool box Ray light trael in traight path alled ray. Index of refration (n) i the ratio of the peed of light () in a auu to the peed of light in

More information

Deterministic Access for DSRC/802.11p Vehicular Safety Communication

Deterministic Access for DSRC/802.11p Vehicular Safety Communication eterminiti Ae for SRC/802.11p Vehiular Safety Communiation Jihene Rezgui, Soumaya Cheraoui, Omar Charoun INTERLAB Reearh Laboratory Univerité de Sherbrooe, Canada {jihene.rezgui, oumaya.heraoui, omar.haroun

More information

Unsupervised Stereoscopic Video Object Segmentation Based on Active Contours and Retrainable Neural Networks

Unsupervised Stereoscopic Video Object Segmentation Based on Active Contours and Retrainable Neural Networks Unsupervised Stereosopi Video Objet Segmentation Based on Ative Contours and Retrainable Neural Networks KLIMIS NTALIANIS, ANASTASIOS DOULAMIS, and NIKOLAOS DOULAMIS National Tehnial University of Athens

More information

Ana C. Andrés del Valle. To cite this version: HAL Id: pastel https://pastel.archives-ouvertes.fr/pastel

Ana C. Andrés del Valle. To cite this version: HAL Id: pastel https://pastel.archives-ouvertes.fr/pastel Analye de mouvement faial dur de image monoulaire ave appliation aux téléommuniation: Couplage de la ompréhenion de l expreion et du uivi de la poe du viage Ana C. André del Valle To ite thi verion: Ana

More information

A Novel Validity Index for Determination of the Optimal Number of Clusters

A Novel Validity Index for Determination of the Optimal Number of Clusters IEICE TRANS. INF. & SYST., VOL.E84 D, NO.2 FEBRUARY 2001 281 LETTER A Novel Validity Index for Determination of the Optimal Number of Clusters Do-Jong KIM, Yong-Woon PARK, and Dong-Jo PARK, Nonmembers

More information

A SIMPLE IMPERATIVE LANGUAGE THE STORE FUNCTION NON-TERMINATING COMMANDS

A SIMPLE IMPERATIVE LANGUAGE THE STORE FUNCTION NON-TERMINATING COMMANDS A SIMPLE IMPERATIVE LANGUAGE Eventually we will preent the emantic of a full-blown language, with declaration, type and looping. However, there are many complication, o we will build up lowly. Our firt

More information

arxiv: v1 [cs.db] 13 Sep 2017

arxiv: v1 [cs.db] 13 Sep 2017 An effiient lustering algorithm from the measure of loal Gaussian distribution Yuan-Yen Tai (Dated: May 27, 2018) In this paper, I will introdue a fast and novel lustering algorithm based on Gaussian distribution

More information

The Minimum Redundancy Maximum Relevance Approach to Building Sparse Support Vector Machines

The Minimum Redundancy Maximum Relevance Approach to Building Sparse Support Vector Machines The Minimum Redundany Maximum Relevane Approah to Building Sparse Support Vetor Mahines Xiaoxing Yang, Ke Tang, and Xin Yao, Nature Inspired Computation and Appliations Laboratory (NICAL), Shool of Computer

More information

Performance of a Robust Filter-based Approach for Contour Detection in Wireless Sensor Networks

Performance of a Robust Filter-based Approach for Contour Detection in Wireless Sensor Networks Performance of a Robut Filter-baed Approach for Contour Detection in Wirele Senor Network Hadi Alati, William A. Armtrong, Jr., and Ai Naipuri Department of Electrical and Computer Engineering The Univerity

More information

Markov Random Fields in Image Segmentation

Markov Random Fields in Image Segmentation Preented at SSIP 2011, Szeged, Hungary Markov Random Field in Image Segmentation Zoltan Kato Image Proceing & Computer Graphic Dept. Univerity of Szeged Hungary Zoltan Kato: Markov Random Field in Image

More information

A DYNAMIC ACCESS CONTROL WITH BINARY KEY-PAIR

A DYNAMIC ACCESS CONTROL WITH BINARY KEY-PAIR Malaysian Journal of Computer Siene, Vol 10 No 1, June 1997, pp 36-41 A DYNAMIC ACCESS CONTROL WITH BINARY KEY-PAIR Md Rafiqul Islam, Harihodin Selamat and Mohd Noor Md Sap Faulty of Computer Siene and

More information

Quadrilaterals. Learning Objectives. Pre-Activity

Quadrilaterals. Learning Objectives. Pre-Activity Section 3.4 Pre-Activity Preparation Quadrilateral Intereting geometric hape and pattern are all around u when we tart looking for them. Examine a row of fencing or the tiling deign at the wimming pool.

More information

Cluster-Based Cumulative Ensembles

Cluster-Based Cumulative Ensembles Cluster-Based Cumulative Ensembles Hanan G. Ayad and Mohamed S. Kamel Pattern Analysis and Mahine Intelligene Lab, Eletrial and Computer Engineering, University of Waterloo, Waterloo, Ontario N2L 3G1,

More information

Exploring the Commonality in Feature Modeling Notations

Exploring the Commonality in Feature Modeling Notations Exploring the Commonality in Feature Modeling Notations Miloslav ŠÍPKA Slovak University of Tehnology Faulty of Informatis and Information Tehnologies Ilkovičova 3, 842 16 Bratislava, Slovakia miloslav.sipka@gmail.om

More information

Mining effective design solutions based on a model-driven approach

Mining effective design solutions based on a model-driven approach ata Mining VI 463 Mining effetive design solutions based on a model-driven approah T. Katsimpa 2, S. Sirmakessis 1,. Tsakalidis 1,2 & G. Tzimas 1,2 1 Researh ademi omputer Tehnology Institute, Hellas 2

More information

Key Terms - MinMin, MaxMin, Sufferage, Task Scheduling, Standard Deviation, Load Balancing.

Key Terms - MinMin, MaxMin, Sufferage, Task Scheduling, Standard Deviation, Load Balancing. Volume 3, Iue 11, November 2013 ISSN: 2277 128X International Journal of Advanced Reearch in Computer Science and Software Engineering Reearch Paper Available online at: www.ijarce.com Tak Aignment in

More information

A User-Attention Based Focus Detection Framework and Its Applications

A User-Attention Based Focus Detection Framework and Its Applications A Uer-Attention Baed Focu Detection Framework and It Application Chia-Chiang Ho, Wen-Huang Cheng, Ting-Jian Pan, Ja-Ling Wu Communication and Multimedia Laboratory, Department of Computer Science and Information

More information

Cross-layer Resource Allocation on Broadband Power Line Based on Novel QoS-priority Scheduling Function in MAC Layer

Cross-layer Resource Allocation on Broadband Power Line Based on Novel QoS-priority Scheduling Function in MAC Layer Communiations and Networ, 2013, 5, 69-73 http://dx.doi.org/10.4236/n.2013.53b2014 Published Online September 2013 (http://www.sirp.org/journal/n) Cross-layer Resoure Alloation on Broadband Power Line Based

More information

Classical Univariate Statistics

Classical Univariate Statistics 1 2 Statitial Modelling and Computing Nik Fieller Department of Probability & Statiti Unierity of Sheffield, UK Claial Uniariate Statiti (& Alternatie) Claial tatitial tet and p-alue Simple imulation method

More information

Web Science and additionality

Web Science and additionality Admin tuff... Lecture 1: EITN01 Web Intelligence and Information Retrieval Meage, lide, handout, lab manual and link: http://www.eit.lth.e/coure/eitn01 Contact: Ander Ardö, Ander.Ardo@eit.lth.e, room:

More information

Learning Convention Propagation in BeerAdvocate Reviews from a etwork Perspective. Abstract

Learning Convention Propagation in BeerAdvocate Reviews from a etwork Perspective. Abstract CS 9 Projet Final Report: Learning Convention Propagation in BeerAdvoate Reviews from a etwork Perspetive Abstrat We look at the way onventions propagate between reviews on the BeerAdvoate dataset, and

More information

CleanUp: Improving Quadrilateral Finite Element Meshes

CleanUp: Improving Quadrilateral Finite Element Meshes CleanUp: Improving Quadrilateral Finite Element Meshes Paul Kinney MD-10 ECC P.O. Box 203 Ford Motor Company Dearborn, MI. 8121 (313) 28-1228 pkinney@ford.om Abstrat: Unless an all quadrilateral (quad)

More information

Routing Definition 4.1

Routing Definition 4.1 4 Routing So far, we have only looked at network without dealing with the iue of how to end information in them from one node to another The problem of ending information in a network i known a routing

More information

A {k, n}-secret Sharing Scheme for Color Images

A {k, n}-secret Sharing Scheme for Color Images A {k, n}-seret Sharing Sheme for Color Images Rastislav Luka, Konstantinos N. Plataniotis, and Anastasios N. Venetsanopoulos The Edward S. Rogers Sr. Dept. of Eletrial and Computer Engineering, University

More information

Detecting Moving Targets in Clutter in Airborne SAR via Keystoning and Multiple Phase Center Interferometry

Detecting Moving Targets in Clutter in Airborne SAR via Keystoning and Multiple Phase Center Interferometry Deteting Moving Targets in Clutter in Airborne SAR via Keystoning and Multiple Phase Center Interferometry D. M. Zasada, P. K. Sanyal The MITRE Corp., 6 Eletroni Parkway, Rome, NY 134 (dmzasada, psanyal)@mitre.org

More information

arxiv: v1 [physics.soc-ph] 17 Oct 2013

arxiv: v1 [physics.soc-ph] 17 Oct 2013 Emergene of Blind Area in Information Sreading arxiv:131707v1 [hyi.o-h] 17 Ot 2013 Zi-Ke Zhang 1,2,, Chu-Xu Zhang 1,3,, Xiao-Pu Han 1,2 and Chuang Liu 1,2 1 Intitute of Information Eonomy, Hangzhou Normal

More information

Semi-Supervised Affinity Propagation with Instance-Level Constraints

Semi-Supervised Affinity Propagation with Instance-Level Constraints Semi-Supervised Affinity Propagation with Instane-Level Constraints Inmar E. Givoni, Brendan J. Frey Probabilisti and Statistial Inferene Group University of Toronto 10 King s College Road, Toronto, Ontario,

More information

Gray Codes for Reflectable Languages

Gray Codes for Reflectable Languages Gray Codes for Refletable Languages Yue Li Joe Sawada Marh 8, 2008 Abstrat We lassify a type of language alled a refletable language. We then develop a generi algorithm that an be used to list all strings

More information

Drawing Lines in 2 Dimensions

Drawing Lines in 2 Dimensions Drawing Line in 2 Dimenion Drawing a traight line (or an arc) between two end point when one i limited to dicrete pixel require a bit of thought. Conider the following line uperimpoed on a 2 dimenional

More information

Kinematic design of a double wishbone type front suspension mechanism using multi-objective optimization

Kinematic design of a double wishbone type front suspension mechanism using multi-objective optimization 5 th utralaian Congre on pplied Mehani, CM 2007 10-12 Deember 2007, Bribane, utralia Kinemati deign of a double wihbone tpe front upenion mehanim uing multi-objetive optimiation J. S. wang 1, S. R. Kim

More information

Detection and Recognition of Non-Occluded Objects using Signature Map

Detection and Recognition of Non-Occluded Objects using Signature Map 6th WSEAS International Conferene on CIRCUITS, SYSTEMS, ELECTRONICS,CONTROL & SIGNAL PROCESSING, Cairo, Egypt, De 9-31, 007 65 Detetion and Reognition of Non-Oluded Objets using Signature Map Sangbum Park,

More information

Keywords Cloud Computing, Service Level Agreements (SLA), CloudSim, Monitoring & Controlling SLA Agent, JADE

Keywords Cloud Computing, Service Level Agreements (SLA), CloudSim, Monitoring & Controlling SLA Agent, JADE Volume 5, Iue 8, Augut 2015 ISSN: 2277 128X International Journal of Advanced Reearch in Computer Science and Software Engineering Reearch Paper Available online at: www.ijarce.com Verification of Agent

More information

Particle Swarm Optimization for the Design of High Diffraction Efficient Holographic Grating

Particle Swarm Optimization for the Design of High Diffraction Efficient Holographic Grating Original Artile Partile Swarm Optimization for the Design of High Diffration Effiient Holographi Grating A.K. Tripathy 1, S.K. Das, M. Sundaray 3 and S.K. Tripathy* 4 1, Department of Computer Siene, Berhampur

More information

Evolutionary Feature Synthesis for Image Databases

Evolutionary Feature Synthesis for Image Databases Evolutionary Feature Synthesis for Image Databases Anlei Dong, Bir Bhanu, Yingqiang Lin Center for Researh in Intelligent Systems University of California, Riverside, California 92521, USA {adong, bhanu,

More information

Self-Adaptive Parent to Mean-Centric Recombination for Real-Parameter Optimization

Self-Adaptive Parent to Mean-Centric Recombination for Real-Parameter Optimization Self-Adaptive Parent to Mean-Centri Reombination for Real-Parameter Optimization Kalyanmoy Deb and Himanshu Jain Department of Mehanial Engineering Indian Institute of Tehnology Kanpur Kanpur, PIN 86 {deb,hjain}@iitk.a.in

More information

Representations and Transformations. Objectives

Representations and Transformations. Objectives Repreentation and Tranformation Objective Derive homogeneou coordinate tranformation matrice Introduce tandard tranformation - Rotation - Tranlation - Scaling - Shear Scalar, Point, Vector Three baic element

More information

Analyzing Hydra Historical Statistics Part 2

Analyzing Hydra Historical Statistics Part 2 Analyzing Hydra Hitorical Statitic Part Fabio Maimo Ottaviani EPV Technologie White paper 5 hnode HSM Hitorical Record The hnode i the hierarchical data torage management node and ha to perform all the

More information

Gray-level histogram. Intensity (grey-level) transformation, or mapping. Use of intensity transformations:

Gray-level histogram. Intensity (grey-level) transformation, or mapping. Use of intensity transformations: Faculty of Informatic Eötvö Loránd Univerity Budapet, Hungary Lecture : Intenity Tranformation Image enhancement by point proceing Spatial domain and frequency domain method Baic Algorithm for Digital

More information

Automatic Physical Design Tuning: Workload as a Sequence Sanjay Agrawal Microsoft Research One Microsoft Way Redmond, WA, USA +1-(425)

Automatic Physical Design Tuning: Workload as a Sequence Sanjay Agrawal Microsoft Research One Microsoft Way Redmond, WA, USA +1-(425) Automati Physial Design Tuning: Workload as a Sequene Sanjay Agrawal Mirosoft Researh One Mirosoft Way Redmond, WA, USA +1-(425) 75-357 sagrawal@mirosoft.om Eri Chu * Computer Sienes Department University

More information

Algorithms, Mechanisms and Procedures for the Computer-aided Project Generation System

Algorithms, Mechanisms and Procedures for the Computer-aided Project Generation System Algorithms, Mehanisms and Proedures for the Computer-aided Projet Generation System Anton O. Butko 1*, Aleksandr P. Briukhovetskii 2, Dmitry E. Grigoriev 2# and Konstantin S. Kalashnikov 3 1 Department

More information

See chapter 8 in the textbook. Dr Muhammad Al Salamah, Industrial Engineering, KFUPM

See chapter 8 in the textbook. Dr Muhammad Al Salamah, Industrial Engineering, KFUPM Goal programming Objective of the topic: Indentify indutrial baed ituation where two or more objective function are required. Write a multi objective function model dla a goal LP Ue weighting um and preemptive

More information

On - Line Path Delay Fault Testing of Omega MINs M. Bellos 1, E. Kalligeros 1, D. Nikolos 1,2 & H. T. Vergos 1,2

On - Line Path Delay Fault Testing of Omega MINs M. Bellos 1, E. Kalligeros 1, D. Nikolos 1,2 & H. T. Vergos 1,2 On - Line Path Delay Fault Testing of Omega MINs M. Bellos, E. Kalligeros, D. Nikolos,2 & H. T. Vergos,2 Dept. of Computer Engineering and Informatis 2 Computer Tehnology Institute University of Patras,

More information

Graph-Based vs Depth-Based Data Representation for Multiview Images

Graph-Based vs Depth-Based Data Representation for Multiview Images Graph-Based vs Depth-Based Data Representation for Multiview Images Thomas Maugey, Antonio Ortega, Pasal Frossard Signal Proessing Laboratory (LTS), Eole Polytehnique Fédérale de Lausanne (EPFL) Email:

More information

Extracting Partition Statistics from Semistructured Data

Extracting Partition Statistics from Semistructured Data Extrating Partition Statistis from Semistrutured Data John N. Wilson Rihard Gourlay Robert Japp Mathias Neumüller Department of Computer and Information Sienes University of Strathlyde, Glasgow, UK {jnw,rsg,rpj,mathias}@is.strath.a.uk

More information

An Edge-based Clustering Algorithm to Detect Social Circles in Ego Networks

An Edge-based Clustering Algorithm to Detect Social Circles in Ego Networks JOURNAL OF COMPUTERS, VOL. 8, NO., OCTOBER 23 2575 An Edge-based Clustering Algorithm to Detet Soial Cirles in Ego Networks Yu Wang Shool of Computer Siene and Tehnology, Xidian University Xi an,77, China

More information

PathRings. Manual. Version 1.0. Yongnan Zhu December 16,

PathRings. Manual. Version 1.0. Yongnan Zhu   December 16, PathRings Version 1.0 Manual Yongnan Zhu E-mail: yongnan@umb.edu Deember 16, 2014-1 - PathRings This tutorial introdues PathRings, the user interfae and the interation. For better to learn, you will need

More information

FUZZY WATERSHED FOR IMAGE SEGMENTATION

FUZZY WATERSHED FOR IMAGE SEGMENTATION FUZZY WATERSHED FOR IMAGE SEGMENTATION Ramón Moreno, Manuel Graña Computational Intelligene Group, Universidad del País Vaso, Spain http://www.ehu.es/winto; {ramon.moreno,manuel.grana}@ehu.es Abstrat The

More information

LinkGuide: Towards a Better Collection of Hyperlinks in a Website Homepage

LinkGuide: Towards a Better Collection of Hyperlinks in a Website Homepage Proceeding of the World Congre on Engineering 2007 Vol I LinkGuide: Toward a Better Collection of Hyperlink in a Webite Homepage A. Ammari and V. Zharkova chool of Informatic, Univerity of Bradford anammari@bradford.ac.uk,

More information

Chapter 2: Introduction to Maple V

Chapter 2: Introduction to Maple V Chapter 2: Introdution to Maple V 2-1 Working with Maple Worksheets Try It! (p. 15) Start a Maple session with an empty worksheet. The name of the worksheet should be Untitled (1). Use one of the standard

More information

A Novel Bit Level Time Series Representation with Implication of Similarity Search and Clustering

A Novel Bit Level Time Series Representation with Implication of Similarity Search and Clustering A Novel Bit Level Time Series Representation with Impliation of Similarity Searh and lustering hotirat Ratanamahatana, Eamonn Keogh, Anthony J. Bagnall 2, and Stefano Lonardi Dept. of omputer Siene & Engineering,

More information

Karen L. Collins. Wesleyan University. Middletown, CT and. Mark Hovey MIT. Cambridge, MA Abstract

Karen L. Collins. Wesleyan University. Middletown, CT and. Mark Hovey MIT. Cambridge, MA Abstract Mot Graph are Edge-Cordial Karen L. Collin Dept. of Mathematic Weleyan Univerity Middletown, CT 6457 and Mark Hovey Dept. of Mathematic MIT Cambridge, MA 239 Abtract We extend the definition of edge-cordial

More information

Advanced Encryption Standard and Modes of Operation

Advanced Encryption Standard and Modes of Operation Advanced Encryption Standard and Mode of Operation G. Bertoni L. Breveglieri Foundation of Cryptography - AES pp. 1 / 50 AES Advanced Encryption Standard (AES) i a ymmetric cryptographic algorithm AES

More information

FOREGROUND OBJECT EXTRACTION USING FUZZY C MEANS WITH BIT-PLANE SLICING AND OPTICAL FLOW

FOREGROUND OBJECT EXTRACTION USING FUZZY C MEANS WITH BIT-PLANE SLICING AND OPTICAL FLOW FOREGROUND OBJECT EXTRACTION USING FUZZY C EANS WITH BIT-PLANE SLICING AND OPTICAL FLOW SIVAGAI., REVATHI.T, JEGANATHAN.L 3 APSG, SCSE, VIT University, Chennai, India JRF, DST, Dehi, India. 3 Professor,

More information

Model Based Approach for Content Based Image Retrievals Based on Fusion and Relevancy Methodology

Model Based Approach for Content Based Image Retrievals Based on Fusion and Relevancy Methodology The International Arab Journal of Information Tehnology, Vol. 12, No. 6, November 15 519 Model Based Approah for Content Based Image Retrievals Based on Fusion and Relevany Methodology Telu Venkata Madhusudhanarao

More information

Is This A Quadrisected Mesh?

Is This A Quadrisected Mesh? I Thi Quadrieted Meh? Gariel Tauin alifornia Intitute of Tehnology Tehnial Report STR 2000008 Deemer 2000 STRT In thi paper we introdue a fat and effiient linear time and pae algorithm to detet and reontrut

More information

What are Cycle-Stealing Systems Good For? A Detailed Performance Model Case Study

What are Cycle-Stealing Systems Good For? A Detailed Performance Model Case Study What are Cyle-Stealing Systems Good For? A Detailed Performane Model Case Study Wayne Kelly and Jiro Sumitomo Queensland University of Tehnology, Australia {w.kelly, j.sumitomo}@qut.edu.au Abstrat The

More information

Building a Compact On-line MRF Recognizer for Large Character Set using Structured Dictionary Representation and Vector Quantization Technique

Building a Compact On-line MRF Recognizer for Large Character Set using Structured Dictionary Representation and Vector Quantization Technique 202 International Conference on Frontier in Handwriting Recognition Building a Compact On-line MRF Recognizer for Large Character Set uing Structured Dictionary Repreentation and Vector Quantization Technique

More information

Video Data and Sonar Data: Real World Data Fusion Example

Video Data and Sonar Data: Real World Data Fusion Example 14th International Conferene on Information Fusion Chiago, Illinois, USA, July 5-8, 2011 Video Data and Sonar Data: Real World Data Fusion Example David W. Krout Applied Physis Lab dkrout@apl.washington.edu

More information

Exploiting Enriched Contextual Information for Mobile App Classification

Exploiting Enriched Contextual Information for Mobile App Classification Exploiting Enrihed Contextual Information for Mobile App Classifiation Hengshu Zhu 1 Huanhuan Cao 2 Enhong Chen 1 Hui Xiong 3 Jilei Tian 2 1 University of Siene and Tehnology of China 2 Nokia Researh Center

More information

xy-monotone path existence queries in a rectilinear environment

xy-monotone path existence queries in a rectilinear environment CCCG 2012, Charlottetown, P.E.I., Augut 8 10, 2012 xy-monotone path exitence querie in a rectilinear environment Gregory Bint Anil Mahehwari Michiel Smid Abtract Given a planar environment coniting of

More information

SLA Adaptation for Service Overlay Networks

SLA Adaptation for Service Overlay Networks SLA Adaptation for Service Overlay Network Con Tran 1, Zbigniew Dziong 1, and Michal Pióro 2 1 Department of Electrical Engineering, École de Technologie Supérieure, Univerity of Quebec, Montréal, Canada

More information

特集 Road Border Recognition Using FIR Images and LIDAR Signal Processing

特集 Road Border Recognition Using FIR Images and LIDAR Signal Processing デンソーテクニカルレビュー Vol. 15 2010 特集 Road Border Reognition Using FIR Images and LIDAR Signal Proessing 高木聖和 バーゼル ファルディ Kiyokazu TAKAGI Basel Fardi ヘンドリック ヴァイゲル Hendrik Weigel ゲルド ヴァニーリック Gerd Wanielik This paper

More information

Approximate logic synthesis for error tolerant applications

Approximate logic synthesis for error tolerant applications Approximate logi synthesis for error tolerant appliations Doohul Shin and Sandeep K. Gupta Eletrial Engineering Department, University of Southern California, Los Angeles, CA 989 {doohuls, sandeep}@us.edu

More information

RAC 2 E: Novel Rendezvous Protocol for Asynchronous Cognitive Radios in Cooperative Environments

RAC 2 E: Novel Rendezvous Protocol for Asynchronous Cognitive Radios in Cooperative Environments 21st Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communiations 1 RAC 2 E: Novel Rendezvous Protool for Asynhronous Cognitive Radios in Cooperative Environments Valentina Pavlovska,

More information

timestamp, if silhouette(x, y) 0 0 if silhouette(x, y) = 0, mhi(x, y) = and mhi(x, y) < timestamp - duration mhi(x, y), else

timestamp, if silhouette(x, y) 0 0 if silhouette(x, y) = 0, mhi(x, y) = and mhi(x, y) < timestamp - duration mhi(x, y), else 3rd International Conferene on Multimedia Tehnolog(ICMT 013) An Effiient Moving Target Traking Strateg Based on OpenCV and CAMShift Theor Dongu Li 1 Abstrat Image movement involved bakground movement and

More information