Optimized Region Competition Algorithm Applied to the Segmentation of Artificial Muscles in Stereoscopic Images

Size: px
Start display at page:

Download "Optimized Region Competition Algorithm Applied to the Segmentation of Artificial Muscles in Stereoscopic Images"

Transcription

1 Vol. 2, No. 3, Page Copyrght 2008, TSI Press Prnted n the USA. All rghts reserved Optmzed Regon Competton Algorthm Appled to the Segmentaton of Artfcal Muscles n Stereoscopc Images Rafael Verdú-Monedero, Juan Morales-Sánchez, and Jorge Larrey-Ruz Department of Informaton Technologes and Communcatons Techncal Unversty of Cartagena {rafael.verdu, juan.morales, jorge.larrey}@upct.es Receved 29 June 2007; revsed 30 November 2007, accepted 25 December 2007 Abstract Ths paper addresses the mplementaton of the Regon Competton algorthm for segmentng stereoscopc vdeo sequences. The segmentaton performed by ths algorthm s an essental stage for the 3D characterzaton of artfcal muscles. Image sequences are acqured by a two-cam computer vson system. The am of ths work s to come up wth optmal and effcent segmentaton of these mages; the nformaton obtaned from the frst segmented frame of the vdeo sequence s used for segmentng the next frame and so on. Redundancy between stereoscopc pars of mages s also used to optmze the segmentaton. In ths paper, the Regon Competton algorthm s descrbed and a specfc mplementaton s detaled. It also shows how to solve partcular problems of stereoscopc vdeo segmentaton. Fnally, results yelded from experments are presented and dscussed. Keywords mage segmentaton, stereo vson, artfcal muscles 1. INTRODUCTION In ths paper, an algorthm specfcally desgned for the segmentaton of stereoscopc mage sequences s descrbed. The mages to be segmented are sequental pctures whch contan artfcal muscles n moton. These devces called artfcal muscles [1] are bult wth polymer conductors provdng the devce wth the ablty to transform electrcal energy nto mechanc work. When the artfcal muscles are exposed to an electrcal ntensty, they can bend or stretch themselves dependng on the sgn of that ntensty. There's a wde range of applcatons for artfcal muscles, from human prosthess to motor systems for small robots, and new ones are appearng every day. Inspred by natural muscles, polymer-made muscles are excted by electrcal ntenstes and ther movement s due to dmensonal varatons of the polymers caused by electrochemcal reactons. Ths stuaton creates the need of a system whch measures small varatons of the artfcal muscles and whch controls ther behavour under dfferent condtons. Observaton and characterzaton of these devces are essental tasks for ts nvestgaton and mprovement. In [2][3] a new method based on actve contours s developed n order to track the moton of the muscle and obtan ts 3D characterzaton. Two dgtal cameras take orthogonal pctures of the muscle n moton. Usng stereoscopc vson technques and dgtal mage processng we get an automatc way to extract the parameters we're lookng for, [4]. Ths method s based on artfcal vson and mage processng. The dfferent stages are:

2 Camera control and stereoscopc mages acquston (see Fgure 1). Image Processng, whch ncludes mage segmentaton (.e., detecton of the muscle n the mages and separaton from the background) and trackng usng a 3D actve contour model [5]. Data mnng: parameters of movement, energy of curvature, etc. Ths work focuses on the mage processng stage, whch s a key stage for the whole process snce t s mpossble to characterze the muscle f the obtaned parameters are naccurate. In ths paper we descrbe the proposed algorthm that carres out the optmzed segmentaton. The startng pont for our segmentaton algorthm s Zhu and Yulle's Regon Competton algorthm [6]. Fgure 1. Stereoscopc magng system for characterzaton of artfcal muscles. On the left, a par of acqured stereo-frames. On the rght, the arrangement of the cameras and the muscle s shown. Segmentaton technques When analysng the objects n mages t becomes essental to dstngush between the objects of nterest (target) and the rest ' (also referred to as the background). Before developng our algorthm, a wde study of several segmentaton technques has been carred out: edge detecton technques, thresholdng, regon growng, actve contours, etc. All of these technques have somethng n common: they formulate some hypotheses about the mage, test features, and make decsons by applyng thresholds explctly or mplctly. The man dfference between these dfferent approaches les n the domans on whch the hypotheses, tests, and decsons are based. Edge based technques only make use of local nformaton and cannot guarantee contnuous closed edge contours. Actve contour models make use only of nformaton along the boundary and requre good ntal estmates to yeld correct convergence. An advantage of regon growng s that t tests the statstcs nsde the regon; however t often generates rregular boundares and small holes. In addton, all of these three methods lack of a global crteron for segmentng the entre mage. In contrast, global optmzaton approaches based on energy functons or Bayesan and MDL (Mnmum Descrpton Length) crtera often have problems when tryng to fnd ther mnma. 186

3 Zhu and Yulle presented n [6] a statstcal framework for mage segmentaton usng a novel algorthm whch they called Regon Competton (RC). It s derved by mnmzng a generalzed Bayes/MDL crteron and combnes the attractve geometrcal features of snake/balloon models and the statstcal technques of regon growng. RC uses a samplng wndow whose sze depends on the sgnal to nose rato. The precson of the boundary locaton depends on the sze of these samplng wndows. Lke many other algorthms, the performance of regon competton depends on the ntal condtons, more precsely on the choce of ntal "seeds. Zhu and Yulle showed a crteron for choosng such seeds and also descrbed an mportant domno effect whereby bad seeds are transformed nto good seeds. RC can be generalzed to multband segmentaton (colour mages) and s able to use several descrptors (as materals texture) to carry out the segmentaton. These features avod the key drawback of many approaches to mage segmentaton whch only look for dscontnutes n ntenstes, and whch can gve msleadng results. The work descrbed n ths paper focuses on gray level mages but the number of descrptors can be ncreased n order to segment more complex mages. The Regon Competton algorthm The goal of mage segmentaton s to partton the mage nto subregons wth homogeneous ntensty (colour or texture) propertes whch wll hopefully correspond to objects or object parts. It s assumed that a regon R s consdered to be homogeneous f ts ntensty values are consstent wth havng been generated by one of a famly of prespecfed probablty dstrbutons P ( I ), where denotes the parameters of the dstrbuton. It s also supposed that the entre mage doman R has been ntally segmented nto M pecewse homogeneous underlyng regons R where =1,2,..., M and R = M R, =1 R R j = 0 f j. Let Γ = R be the R boundary of regon, where we defne the drecton of to be counter-clockwse. Let M = 1 R Γ = Γ be the edges or segmentaton boundares of the entre mage. Now consder an MDL crteron (a global energy functonal). Ths gves: E[ Γ, ] = M μ { ds ({ I( x, y) : ( x, y) R } ) + λ} 2 log P, = 1 R (1) where the frst term wthn the braces s the length of the boundary curve R, for regon R. We smply assume that the code length s proportonal to the curve length, where μ s the code length for unt arc length. Snce all edge segments are shared by two adjacent regons, we dvde the frst term by a factor of 2. The second term s the sum of the cost for codng the ntensty of every pxel ( x, y) nsde regon R, accordng to a dstrbuton P( { I( x, y) : ( x, y) R } ). λ s the output of a cost functon whch s related to the code length needed to descrbe the dstrbuton and code system for regon R, and we smply assume that λ s common to all regons. Because the energy E n (1) depends on two groups of varables (the segmentaton Γ and the parameters ), Zhu and Yulle proposed a greedy algorthm whch conssts of two alternatng stages. The frst stage locally mnmzes the energy wth the number of regons fxed. It proceeds by teratng two steps both of whch cause the energy to decrease. The second stage merges regons provded ths decreases the energy. The frst stage ncludes two steps. In the frst step, we fx Γ and we solve for to mnmze the descrpton cost for each regon. In other words, parameters are estmated by maxmzng the condtonal probabltes. In the second step, we fx the 187

4 parameters of the dstrbuton and do steepest descent wth respect to Γ. Both steps n ths frst stage of the algorthm cause the energy functon to decrease. In addton, the functon s bounded below and so the algorthm s guaranteed to converge to a local mnmum. Ths two step process, however, does not allow us to alter the number of regons. Thus a second stage s added where adjacent regons are merged f ths causes the energy to decrease. Ths s followed by the two step teraton stage agan, and so on. Overall each operaton reduces the energy and so a local mnmum s reached. The whole process, as descrbed n [6], goes as follows: 1. Intalze the segmentaton; put N seeds randomly across the mage; all background s treated as a sngle regon wth unform probablty dstrbuton. 2. Fx the boundary Γ ; compute the parameters by maxmzng P( I : ). 3. Fx, move the boundary Γ by mnmzng the energy functon. When two seed regons meet, an edge s formed at ther common boundary, and then these two regons compete along ths boundary. 4. Execute step 2, 3 teratvely untl the moton of boundary Γ converges. Then go to step If there s any background regon not occuped by any seed regons, then put a new seed n the background, and go to step 2; else go to step Merge two adjacent regons so that the mergng causes the largest energy decrease; go to step 2. If no merge can decrease the energy, then go to step Stop. 2. RC BASED SEGMENTATION OF STEREOSCOPIC SEQUENCES OF ARTIFICIAL MUSCLES We descrbe n ths Secton our adaptaton of RC algorthm for artfcal muscles segmentaton. As shown n the ntroducton of ths paper, our segmentaton work s a key stage n the method desgned to obtan a characterzaton of artfcal muscles. The nputs for our applcaton are pars of orthogonal pctures of an artfcal muscle n moton (see Fgure 1). Our algorthm s responsble of segmentng these mages along the stereoscopc sequence. The output generated wll be used by actve contours to obtan the parameters that characterze the muscle movements. In the followng, we expose a general overvew of how our algorthm works, and we show how t deals wth mportant ssues such as the ntal segmentaton, the use of redundancy n stereoscopc pars of frames and the trackng of the target along the sequence. General overvew of the algorthm operaton The easest way to understand our algorthm s to descrbe ts operaton from the frst par of frames acqured. Let t) be the frame captured n the XZ plane and let B(t) be the frame captured n the YZ plane at the same nstantt. Fgure 2 shows a sequence of three stereo frames. As shown n the fgure, dfferences between consecutve frames are small. Ths wll help the target trackng stage later. Also, both cameras are calbrated so that the relatve poston of the muscle s the same n both frames t) and B(t). The presented applcaton operates wth pars of frames t), B(t) smultaneously. The frst par 0), B(0) s segmented wthout any a pror nformaton, followng the algorthm descrbed n Secton 1.2. The RC 188

5 Input Intal segmentaton Output t=0 Trackng Output t=1 Trackng Output t=2 Fgure 2. Sequence segmentaton flow. Whte arrows show nformaton transfer. 0) algorthm s appled to wth N seeds randomly put n the mage. The process follows the algorthm descrbed n the prevous secton untl the segmentaton Γ s complete. Once the ntal segmentaton of 0) s done (.e., the whole mage I( x, y) s parttoned nto N regons R ), the regon of nterest (the artfcal muscle) can be chosen from the descrptors of each regon or t s possble that the user can be prompted to choose the regon of nterest (t s useful n other applcatons wth more complex mages to segment). Now the algorthm knows what to look for and then t uses that nformaton to segment B(0) much faster. Ths ntal segmentaton could be accelerated by usng the knowledge we already have about the mages, but ths way we ensure that the algorthm wll work wth mages of any knd. At ths pont, we have the segmentaton Γ of both ntal frames A (0) and B(0). From now on, the algorthm wll work n a statonary mode untl the end of the whole sequence. In ths stage, we do have nformaton about prevous frames and we can use t to segment consecutve mages much faster. In ths statonary mode, the RC algorthm s not fully used as descrbed n the prevous secton. Instead of that, the number of regons s lmted to target and background and the mergng step almost dsappears, so the computatonal cost s reduced. Target trackng through the stereoscopc vdeo s done by change detecton n consecutve frames. Frame dfference s calculated between frames t) and A ( t 1) and the regon competton algorthm only takes care about the pxels that have changed. 189

6 Intal segmentaton As explaned n the prevous secton, segmentaton of the frst frame 0) s carred out by the RC algorthm descrbed before. Ths frst segmentaton s mportant because the next frames ( B(0), 1) ) are segmented much faster thanks to the nformaton obtaned. Moreover, f the muscle segmentaton n the frst frame s wrong our applcaton wll probably not be able to track ts movements. In ths secton we explan some mportant ssues related to ths part of the process. Frst step s to put N seeds randomly across the mage (see Fgure 3(b)). The sze of these ntal seeds s 1 pxel but the ntal descrptors,.e. the parameters that characterze the dstrbuton, are calculated wth a 5 5 pxel samplng wndow. Ths s to ensure that ntal descrptors do not correspond to nosy peaks n the mage. If one ntal seed falls nto one of these peaks, the statstcal forces would not let t grow as neghbour pxels do not pass the lkelhood test. On the other hand, the samplng wndow cannot be too bg because f t covers a sgnfcant edge t may be gnored. In fact, the optmal sze of the samplng wndow depends on the mage sgnal-to-nose rato, n [6] a dscusson about seeds and samplng wndows can be found. Now we have an approxmaton of the fnal segmentaton Γ 0). Ths s a m n matrx of labels, whch s the same sze of the frame. Ths matrx has zeros n every poston except n those N pxels wth ntal seeds R. Every seed regon has an assocated descrptor whch s used to control ts growng., Next step s to make those N regons grow. Ths s done by morphologc operators that dlate the ntal regons wth a 3 3 pxels structurng element. New pxels are tested aganst the regon descrptors and only become part of the regon f they ft the dstrbuton P I ). Usng gray level ( (a) (b) (c) (d) (e) (f) (g) (h) Fgure 3. Intal segmentaton 0) wth N = 8 random seeds (a) Frame 0). (b) Segmentaton Γ 0) after the frst growng teraton. (c), (d), (e), (f), (g) show Γ 0) evoluton every 10 teratons. (h) Fnal segmentaton Γ 0) after 160 teratons. 190

7 mages and one descrptor per regon, ths decson s taken by comparng the pxel value wth the descrptors, whch are, n ths case, the mean of the pxels n the regon. If the dfference s not hgher than the growng threshold T g, the new pxel s fnally added. Descrptors are recalculated n every growng teraton to keep them representatve of ther regons R. When the values of Γ 0) converge (.e., regons cannot grow anymore) the algorthm looks for pxels wthout any regon assgned and, f there s any, t puts a new seed and the prevous steps are repeated. Once every pxel n the mage has been labelled, the next step s to merge regons f ths causes the energy to decrease, that s, regons wth smlar descrptors are merged usng a new threshold called T m (see Fgure 3 (g) and (h)). If the values chosen for Tg and Tm are too tght, regons wll have problems to grow and new seeds wll be added to complete Γ 0). Ths means more regons, more descrptors and much more processng tme to obtan an over-segmented mage as fnal result. On the other hand, f Tg and Tm are too relaxed our algorthm wll be less senstve to detals and we may loose some regons of nterest. However, most artfcal muscle pctures have a good contrast between target and background, and by knowng the descrptors of prevous segmentatons, t makes easy to tune both thresholds properly (common values are T =50 and T =30). g m Redundancy between stereoscopc pars of mages As stated before, our applcaton only uses the nformaton obtaned from Γ t) to segment B(t) when t = 0. In other words, nformaton transfer between stereoscopc pars only occurs for the frst par of mages (see Fgure 2). It s possble to mplement ths transfer for every frame and t may be useful for mages wth dfferent characterstcs. However, experments carred out wth artfcal muscle sequences do not show any mprovement. In ths secton we descrbe how we use the nformaton obtaned from Γ t) to speed up the segmentaton of B(t). Assumng that Γ t) s correct, the nformaton we can use to segment B(t) s: Muscle poston t ) R n frame A (t). t) nterest Descrptor nterest of the muscle regon t) Rnterest n A (t). We use the muscle poston n t) to delmt the range of pxels where the algorthm wll search for the muscle n B(t). Cameras are calbrated so that both vews of the muscle are n the same relatve poston and scale. Therefore we can assume that muscle coordnates n z axs are the same n t) and B(t). Pxels lyng out of ths range are labelled as background (see Fgure 4(c)). t) The value of descrptor nterest s used to fnd the y axs coordnates of the muscle n B(t). The algorthm looks for pxels wth values t ) smlar to nterest and labels the rest as background (see Fgure 4(d)). At ths pont, the segmentaton of frame B(t) starts by applyng the RC algorthm. However, as more than 90% of the mage s already segmented (labeled as background) there s no competton between regons except n some specal cases. Part of the target gets occluded when the muscle's end reaches the water surface. In ths stuaton our applcaton needs the regon competton strength to complete the segmentaton correctly. Once we have delmted the muscle poston n B(t), the segmentaton starts by puttng one seed at the centre of the delmted zone and then we let t grow untl the segmentaton s fnshed. More 191

8 (a) (b) (c) (d) (e) Fgure 4. Informaton transfer between stereoscopc pars. (a) Frame A (t). (b) Frame B(t). (c) Segmentaton Γ. (d) Search range n B(t). One seed s put n the z coordnate of the t) t) t ) gravty centre of R, y coordnate s found by lookng for smlar pxels to. (e) nterest shows the regons growng. nterest (a) (b) (c) Fgure 5. Estmated target poston usng prevous knowledge. (a) Frame B ( t 1). (b) Target estmaton dlatng prevous segmentaton. (c) Frame B(t). specfcally, z axs (.e., vertcal) coordnate of the ntal seed s found by calculatng t) R gravty centre. The y axs (.e., nterest horzontal) coordnate s chosen by fndng the t ) most smlar pxel to nterest. Intal seed poston s carefully chosen because we do not want regons to compete when t s not necessary. Vdeo Segmentaton: Muscle Trackng In ths secton t s shown how t) and B(t) frames are segmented by usng the segmentaton of prevous frames. The followng explanaton s referred to YZ plane frames ( B(t) and B( t 1) ), but segmentaton of t) frames s analogous. The technque appled to track the muscle through the vdeo sequence s based on change detecton between consecutve frames. These changes are usually detected by calculatng the dfference I(t) I(t 1) between consecutve frames. In ths case, we use the knowledge that we already have about the mage and ts movement to optmze the segmentaton. As the muscle speed s lmted, we assume that ts poston wll be only slghtly dfferent n B(t). Therefore, we can estmate the muscle poston n B (t) f we know Γ B ( t 1). In order to do ths, the muscle poston n B ( t 1) s dlated morphologcally whch gves us a lmted space where the muscle mght be n B(t) (see Fgure 5 (b)). Once more, most pxels n B(t) are already labelled as background before startng RC algorthm. Now we apply the change detecton technque by calculatng the dfference between B ( t 1) and B(t) only n the estmated zone. The processng tme s reduced 192

9 ths way because most pxels of both frames are gnored. Besdes, we prevent the algorthm from beng dstracted by slght changes s water surface. Next step conssts n fndng the optmal poston to place our ntal seed. We use the B( t 1) nterest descrptor to test the pxels not excluded yet by our estmaton. A new seed s put at the most smlar pxel found n B(t) and t starts growng (see Fgure 6 (b)). In the RC algorthm all regons grow alternatvely to ensure a far ' competton but at ths pont we let the regon of nterest grow frst. Ths s a way to avod problems wth possble changes n llumnaton (due to muscle movement) that may mslead our algorthm. Once the muscle regon s grown, the background regon starts growng untl segmentaton of B (t) s fnshed. Experments show that the trackng process s remarkably accelerated thanks to ths technque because regon competton only occurs n specal cases lke the occluson problem mentoned before. Nevertheless, the performed segmentaton s correct n every case. 3. SEGMENTATION RESULTS Several tests have been run wth dfferent sequences and dfferent types of artfcal muscles and the results yelded are satsfactory. The obtaned segmentatons represent fathfully the muscle contour and are sutable to extract ts parameters. Fgure 7. Fgure 8 and Fgure 9 show the results obtaned n the frst frames of a 60 pars sequences of pxels. In Fgure 7 the muscle has been loaded wth a pece of metal n order to test ts strength. XZ plane segmentaton s correct as t does not add the metal rng to the target regon. Notce the changes n lghtng between frames t) and B(t) and the presence of hard nose. Results obtaned are also satsfactory n ths stuaton. 4. CONCLUSIONS Ths paper shows the applcaton developed for segmentng stereoscopc vdeo sequences based on the regon competton algorthm. The orgnal RC algorthm has been tuned and optmzed for artfcal muscle vdeo segmentaton n gray level mages and results yelded are satsfactory. Parameters of muscle behavour can be obtaned from the output of our algorthm even n adverse stuatons of nose and lghtng. The process can be run wthout any human partcpaton after thresholds are adjusted. Redundancy between stereoscopc pars of mages has been used to optmze the segmentaton wth satsfactory results. Muscle trackng has also been optmzed by estmatng target poston n consecutve frames. As a future work t could be nterestng to extend the method by addng some topologcal model to ensure coherent segmentatons n both planes. 5. ACKNOWLEDGEMENTS Ths work s partally supported by Mnstero de Educacón y Cenca, under grant TEC /TCM, and by Fundacón Séneca, project 03122/PI/05. (a) (b) (c) Fgure 6. (a) Change detecton between B (t) and B ( t 1). (b) Intal seed s placed usng and t starts growng n frst place. (c) Fnal segmentaton. B( t 1) nterest Γ B(t) 193

10 (a) A (0) (b) B (0) (c) A (1) (d) B(1) (e) A (2) (f) B (2) (g) A (3) (h) B(3) Fgure 7. Stereoscopc vdeo segmentaton. The muscle s loaded wth a metal rng. (a) A (0) (b) B (0) (c) A (3) (d) B(3) (e) A (6) (f) B (6) (g) A (9) (h) B(9) Fgure 8. Stereoscopc vdeo segmentaton. Intal stereo frames (a) and (b) and stereo frames 6 and

11 (a) A (0) (b) B (0) (c) A (3) (d) B(3) (e) A (6) (f) B (6) (g) A (9) (h) B(9) Fgure 9. Stereoscopc vdeo segmentaton. Intal stereo frames (a) and (b) and stereo frames 6 and 9. REFERENCES [5] Lang, J., McInerney, T., Terzopoulos, D. Unted snakes. Proc. Int. Conf. Computer Vson, pp , [1] Otero, T.F., Sansñena, J.M. Blayer dmensons and movement of artfcal muscles. Boelectrochem. Boenergetcs, 47, [2] Verdú, R., Morales, J., Fernandez-Romero, A.J., Cortés, M.T., Otero, T.F., Weruaga, L. Mechancal characterzaton of artfcal muscles wth computer vson. Proc. of SPIE Annual Int. Symposum on Smart Structures and Materals, 2002 [3] Verdú, R., Berenguer, R., Morales, J., Vázquez, G., Otero, T. F., Weruaga, L. 3D mechancal characterzaton of artfcal muscles wth stereoscopc computer vson and actve contours. Proc. of IEEE Int. Conf. on Image Processng, [4] Verdú, R., Berenguer, R., Morales, J., Vázquez, G., Otero, T., Weruaga, L. Mechancal characterzaton of the lfe cycle of artfcal muscles through stereoscopc computer vson and actve contours. IEEE Int. Conf. on Image Processng, 2005 [6] Zhu, S.C., Yulle, A.L. Regon competton: Unfyng snakes, regon growng, and bayes/mdl for multband mage segmentaton. IEEE Trans. Pattern Anal. Machne Intell, 18, 9, pp , AUTHOR INFORMATION Rafael Verdu receved the MS degree n telecommuncatons engneerng n 2000 from the Techncal Unversty of Valenca (UPV), and the PhD n the same feld n 2005 from the Techncal Unversty of Cartagena (UPCT). He s currently a researcher and assstant teacher n sgnal and communcatons theory at UPCT. Hs techncal nterests are modelng and nverse problems n mage analyss. 195

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

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

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

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

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

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

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

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

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

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

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

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

Active Contours/Snakes

Active Contours/Snakes Actve Contours/Snakes Erkut Erdem Acknowledgement: The sldes are adapted from the sldes prepared by K. Grauman of Unversty of Texas at Austn Fttng: Edges vs. boundares Edges useful sgnal to ndcate occludng

More information

Parallel matrix-vector multiplication

Parallel matrix-vector multiplication Appendx A Parallel matrx-vector multplcaton The reduced transton matrx of the three-dmensonal cage model for gel electrophoress, descrbed n secton 3.2, becomes excessvely large for polymer lengths more

More information

The Codesign Challenge

The Codesign Challenge ECE 4530 Codesgn Challenge Fall 2007 Hardware/Software Codesgn The Codesgn Challenge Objectves In the codesgn challenge, your task s to accelerate a gven software reference mplementaton as fast as possble.

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

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

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

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

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

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

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

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

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

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Proceedngs of the Wnter Smulaton Conference M E Kuhl, N M Steger, F B Armstrong, and J A Jones, eds A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Mark W Brantley Chun-Hung

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

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

12. Segmentation. Computer Engineering, i Sejong University. Dongil Han

12. Segmentation. Computer Engineering, i Sejong University. Dongil Han Computer Vson 1. Segmentaton Computer Engneerng, Sejong Unversty Dongl Han Image Segmentaton t Image segmentaton Subdvdes an mage nto ts consttuent regons or objects - After an mage has been segmented,

More information

Fitting: Deformable contours April 26 th, 2018

Fitting: Deformable contours April 26 th, 2018 4/6/08 Fttng: Deformable contours Aprl 6 th, 08 Yong Jae Lee UC Davs Recap so far: Groupng and Fttng Goal: move from array of pxel values (or flter outputs) to a collecton of regons, objects, and shapes.

More information

TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z.

TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z. TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS Muradalyev AZ Azerbajan Scentfc-Research and Desgn-Prospectng Insttute of Energetc AZ1012, Ave HZardab-94 E-mal:aydn_murad@yahoocom Importance of

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

Outline. Self-Organizing Maps (SOM) US Hebbian Learning, Cntd. The learning rule is Hebbian like:

Outline. Self-Organizing Maps (SOM) US Hebbian Learning, Cntd. The learning rule is Hebbian like: Self-Organzng Maps (SOM) Turgay İBRİKÇİ, PhD. Outlne Introducton Structures of SOM SOM Archtecture Neghborhoods SOM Algorthm Examples Summary 1 2 Unsupervsed Hebban Learnng US Hebban Learnng, Cntd 3 A

More information

Efficient Video Coding with R-D Constrained Quadtree Segmentation

Efficient Video Coding with R-D Constrained Quadtree Segmentation Publshed on Pcture Codng Symposum 1999, March 1999 Effcent Vdeo Codng wth R-D Constraned Quadtree Segmentaton Cha-Wen Ln Computer and Communcaton Research Labs Industral Technology Research Insttute Hsnchu,

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

Image Representation & Visualization Basic Imaging Algorithms Shape Representation and Analysis. outline

Image Representation & Visualization Basic Imaging Algorithms Shape Representation and Analysis. outline mage Vsualzaton mage Vsualzaton mage Representaton & Vsualzaton Basc magng Algorthms Shape Representaton and Analyss outlne mage Representaton & Vsualzaton Basc magng Algorthms Shape Representaton and

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

Programming in Fortran 90 : 2017/2018

Programming in Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Exercse 1 : Evaluaton of functon dependng on nput Wrte a program who evaluate the functon f (x,y) for any two user specfed values

More information

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation 17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 An Iteratve Soluton Approach to Process Plant Layout usng Mxed

More information

Wishing you all a Total Quality New Year!

Wishing you all a Total Quality New Year! Total Qualty Management and Sx Sgma Post Graduate Program 214-15 Sesson 4 Vnay Kumar Kalakband Assstant Professor Operatons & Systems Area 1 Wshng you all a Total Qualty New Year! Hope you acheve Sx sgma

More information

Corner-Based Image Alignment using Pyramid Structure with Gradient Vector Similarity

Corner-Based Image Alignment using Pyramid Structure with Gradient Vector Similarity Journal of Sgnal and Informaton Processng, 013, 4, 114-119 do:10.436/jsp.013.43b00 Publshed Onlne August 013 (http://www.scrp.org/journal/jsp) Corner-Based Image Algnment usng Pyramd Structure wth Gradent

More information

NAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics

NAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics Introducton G10 NAG Fortran Lbrary Chapter Introducton G10 Smoothng n Statstcs Contents 1 Scope of the Chapter... 2 2 Background to the Problems... 2 2.1 Smoothng Methods... 2 2.2 Smoothng Splnes and Regresson

More information

Real-time Motion Capture System Using One Video Camera Based on Color and Edge Distribution

Real-time Motion Capture System Using One Video Camera Based on Color and Edge Distribution Real-tme Moton Capture System Usng One Vdeo Camera Based on Color and Edge Dstrbuton YOSHIAKI AKAZAWA, YOSHIHIRO OKADA, AND KOICHI NIIJIMA Graduate School of Informaton Scence and Electrcal Engneerng,

More information

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009.

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009. Farrukh Jabeen Algorthms 51 Assgnment #2 Due Date: June 15, 29. Assgnment # 2 Chapter 3 Dscrete Fourer Transforms Implement the FFT for the DFT. Descrbed n sectons 3.1 and 3.2. Delverables: 1. Concse descrpton

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

Optimizing Document Scoring for Query Retrieval

Optimizing Document Scoring for Query Retrieval Optmzng Document Scorng for Query Retreval Brent Ellwen baellwe@cs.stanford.edu Abstract The goal of ths project was to automate the process of tunng a document query engne. Specfcally, I used machne learnng

More information

Smoothing Spline ANOVA for variable screening

Smoothing Spline ANOVA for variable screening Smoothng Splne ANOVA for varable screenng a useful tool for metamodels tranng and mult-objectve optmzaton L. Rcco, E. Rgon, A. Turco Outlne RSM Introducton Possble couplng Test case MOO MOO wth Game Theory

More information

Lecture #15 Lecture Notes

Lecture #15 Lecture Notes Lecture #15 Lecture Notes The ocean water column s very much a 3-D spatal entt and we need to represent that structure n an economcal way to deal wth t n calculatons. We wll dscuss one way to do so, emprcal

More information

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters Proper Choce of Data Used for the Estmaton of Datum Transformaton Parameters Hakan S. KUTOGLU, Turkey Key words: Coordnate systems; transformaton; estmaton, relablty. SUMMARY Advances n technologes and

More information

S1 Note. Basis functions.

S1 Note. Basis functions. S1 Note. Bass functons. Contents Types of bass functons...1 The Fourer bass...2 B-splne bass...3 Power and type I error rates wth dfferent numbers of bass functons...4 Table S1. Smulaton results of type

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

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

MOTION PANORAMA CONSTRUCTION FROM STREAMING VIDEO FOR POWER- CONSTRAINED MOBILE MULTIMEDIA ENVIRONMENTS XUNYU PAN

MOTION PANORAMA CONSTRUCTION FROM STREAMING VIDEO FOR POWER- CONSTRAINED MOBILE MULTIMEDIA ENVIRONMENTS XUNYU PAN MOTION PANORAMA CONSTRUCTION FROM STREAMING VIDEO FOR POWER- CONSTRAINED MOBILE MULTIMEDIA ENVIRONMENTS by XUNYU PAN (Under the Drecton of Suchendra M. Bhandarkar) ABSTRACT In modern tmes, more and more

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

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

GSLM Operations Research II Fall 13/14

GSLM Operations Research II Fall 13/14 GSLM 58 Operatons Research II Fall /4 6. Separable Programmng Consder a general NLP mn f(x) s.t. g j (x) b j j =. m. Defnton 6.. The NLP s a separable program f ts objectve functon and all constrants are

More information

A B-Snake Model Using Statistical and Geometric Information - Applications to Medical Images

A B-Snake Model Using Statistical and Geometric Information - Applications to Medical Images A B-Snake Model Usng Statstcal and Geometrc Informaton - Applcatons to Medcal Images Yue Wang, Eam Khwang Teoh and Dnggang Shen 2 School of Electrcal and Electronc Engneerng, Nanyang Technologcal Unversty

More information

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes SPH3UW Unt 7.3 Sphercal Concave Mrrors Page 1 of 1 Notes Physcs Tool box Concave Mrror If the reflectng surface takes place on the nner surface of the sphercal shape so that the centre of the mrror bulges

More information

EXTENDED BIC CRITERION FOR MODEL SELECTION

EXTENDED BIC CRITERION FOR MODEL SELECTION IDIAP RESEARCH REPORT EXTEDED BIC CRITERIO FOR ODEL SELECTIO Itshak Lapdot Andrew orrs IDIAP-RR-0-4 Dalle olle Insttute for Perceptual Artfcal Intellgence P.O.Box 59 artgny Valas Swtzerland phone +4 7

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

A Robust Method for Estimating the Fundamental Matrix

A Robust Method for Estimating the Fundamental Matrix Proc. VIIth Dgtal Image Computng: Technques and Applcatons, Sun C., Talbot H., Ourseln S. and Adraansen T. (Eds.), 0- Dec. 003, Sydney A Robust Method for Estmatng the Fundamental Matrx C.L. Feng and Y.S.

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

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

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

Efficient Content Representation in MPEG Video Databases

Efficient Content Representation in MPEG Video Databases Effcent Content Representaton n MPEG Vdeo Databases Yanns S. Avrths, Nkolaos D. Doulams, Anastasos D. Doulams and Stefanos D. Kollas Department of Electrcal and Computer Engneerng Natonal Techncal Unversty

More information

Nonlocal Mumford-Shah Model for Image Segmentation

Nonlocal Mumford-Shah Model for Image Segmentation for Image Segmentaton 1 College of Informaton Engneerng, Qngdao Unversty, Qngdao, 266000,Chna E-mal:ccluxaoq@163.com ebo e 23 College of Informaton Engneerng, Qngdao Unversty, Qngdao, 266000,Chna E-mal:

More information

Image Segmentation with Simultaneous Illumination and Reflectance Estimation: An Energy Minimization Approach

Image Segmentation with Simultaneous Illumination and Reflectance Estimation: An Energy Minimization Approach Image Segmentaton wth Smultaneous Illumnaton and Reflectance Estmaton: An Energy Mnmzaton Approach Chunmng L 1, Fang L 2, Chu-Yen Kao 3, Chenyang Xu 4 1 Vanderblt Unversty Insttute of Imagng Scence, ashvlle,

More information

Dynamic Camera Assignment and Handoff

Dynamic Camera Assignment and Handoff 12 Dynamc Camera Assgnment and Handoff Br Bhanu and Ymng L 12.1 Introducton...338 12.2 Techncal Approach...339 12.2.1 Motvaton and Problem Formulaton...339 12.2.2 Game Theoretc Framework...339 12.2.2.1

More information

A NEW FUZZY C-MEANS BASED SEGMENTATION STRATEGY. APPLICATIONS TO LIP REGION IDENTIFICATION

A NEW FUZZY C-MEANS BASED SEGMENTATION STRATEGY. APPLICATIONS TO LIP REGION IDENTIFICATION A NEW FUZZY C-MEANS BASED SEGMENTATION STRATEGY. APPLICATIONS TO LIP REGION IDENTIFICATION Mhaela Gordan *, Constantne Kotropoulos **, Apostolos Georgaks **, Ioanns Ptas ** * Bass of Electroncs Department,

More information

A Gradient Difference based Technique for Video Text Detection

A Gradient Difference based Technique for Video Text Detection A Gradent Dfference based Technque for Vdeo Text Detecton Palaahnakote Shvakumara, Trung Quy Phan and Chew Lm Tan School of Computng, Natonal Unversty of Sngapore {shva, phanquyt, tancl }@comp.nus.edu.sg

More information

Circuit Analysis I (ENGR 2405) Chapter 3 Method of Analysis Nodal(KCL) and Mesh(KVL)

Circuit Analysis I (ENGR 2405) Chapter 3 Method of Analysis Nodal(KCL) and Mesh(KVL) Crcut Analyss I (ENG 405) Chapter Method of Analyss Nodal(KCL) and Mesh(KVL) Nodal Analyss If nstead of focusng on the oltages of the crcut elements, one looks at the oltages at the nodes of the crcut,

More information

Range images. Range image registration. Examples of sampling patterns. Range images and range surfaces

Range images. Range image registration. Examples of sampling patterns. Range images and range surfaces Range mages For many structured lght scanners, the range data forms a hghly regular pattern known as a range mage. he samplng pattern s determned by the specfc scanner. Range mage regstraton 1 Examples

More information

Quality Improvement Algorithm for Tetrahedral Mesh Based on Optimal Delaunay Triangulation

Quality Improvement Algorithm for Tetrahedral Mesh Based on Optimal Delaunay Triangulation Intellgent Informaton Management, 013, 5, 191-195 Publshed Onlne November 013 (http://www.scrp.org/journal/m) http://dx.do.org/10.36/m.013.5601 Qualty Improvement Algorthm for Tetrahedral Mesh Based on

More information

Visual Curvature. 1. Introduction. y C. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), June 2007

Visual Curvature. 1. Introduction. y C. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), June 2007 IEEE onf. on omputer Vson and Pattern Recognton (VPR June 7 Vsual urvature HaRong Lu, Longn Jan Lateck, WenYu Lu, Xang Ba HuaZhong Unversty of Scence and Technology, P.R. hna Temple Unversty, US lhrbss@gmal.com,

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

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

Geodesic Active Regions for Supervised Texture Segmentation

Geodesic Active Regions for Supervised Texture Segmentation Geodesc Actve egons for Supervsed Texture Segmentaton Nkos Paragos achd Derche INIA BP 9, 00, oute des Lucoles 0690 Sopha Antpols Cedex, France e-mal: {nparago,der}@sophanrafr Abstract Ths paper presents

More information

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements Module 3: Element Propertes Lecture : Lagrange and Serendpty Elements 5 In last lecture note, the nterpolaton functons are derved on the bass of assumed polynomal from Pascal s trangle for the fled varable.

More information

A Gradient Difference based Technique for Video Text Detection

A Gradient Difference based Technique for Video Text Detection 2009 10th Internatonal Conference on Document Analyss and Recognton A Gradent Dfference based Technque for Vdeo Text Detecton Palaahnakote Shvakumara, Trung Quy Phan and Chew Lm Tan School of Computng,

More information

Using Fuzzy Logic to Enhance the Large Size Remote Sensing Images

Using Fuzzy Logic to Enhance the Large Size Remote Sensing Images Internatonal Journal of Informaton and Electroncs Engneerng Vol. 5 No. 6 November 015 Usng Fuzzy Logc to Enhance the Large Sze Remote Sensng Images Trung Nguyen Tu Huy Ngo Hoang and Thoa Vu Van Abstract

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

Multiple Frame Motion Inference Using Belief Propagation

Multiple Frame Motion Inference Using Belief Propagation Multple Frame Moton Inference Usng Belef Propagaton Jang Gao Janbo Sh The Robotcs Insttute Department of Computer and Informaton Scence Carnege Mellon Unversty Unversty of Pennsylvana Pttsburgh, PA 53

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

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz Compler Desgn Sprng 2014 Regster Allocaton Sample Exercses and Solutons Prof. Pedro C. Dnz USC / Informaton Scences Insttute 4676 Admralty Way, Sute 1001 Marna del Rey, Calforna 90292 pedro@s.edu Regster

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

A Modified Median Filter for the Removal of Impulse Noise Based on the Support Vector Machines

A Modified Median Filter for the Removal of Impulse Noise Based on the Support Vector Machines A Modfed Medan Flter for the Removal of Impulse Nose Based on the Support Vector Machnes H. GOMEZ-MORENO, S. MALDONADO-BASCON, F. LOPEZ-FERRERAS, M. UTRILLA- MANSO AND P. GIL-JIMENEZ Departamento de Teoría

More information

Sequential search. Building Java Programs Chapter 13. Sequential search. Sequential search

Sequential search. Building Java Programs Chapter 13. Sequential search. Sequential search Sequental search Buldng Java Programs Chapter 13 Searchng and Sortng sequental search: Locates a target value n an array/lst by examnng each element from start to fnsh. How many elements wll t need to

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

AMath 483/583 Lecture 21 May 13, Notes: Notes: Jacobi iteration. Notes: Jacobi with OpenMP coarse grain

AMath 483/583 Lecture 21 May 13, Notes: Notes: Jacobi iteration. Notes: Jacobi with OpenMP coarse grain AMath 483/583 Lecture 21 May 13, 2011 Today: OpenMP and MPI versons of Jacob teraton Gauss-Sedel and SOR teratve methods Next week: More MPI Debuggng and totalvew GPU computng Read: Class notes and references

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

User Authentication Based On Behavioral Mouse Dynamics Biometrics

User Authentication Based On Behavioral Mouse Dynamics Biometrics User Authentcaton Based On Behavoral Mouse Dynamcs Bometrcs Chee-Hyung Yoon Danel Donghyun Km Department of Computer Scence Department of Computer Scence Stanford Unversty Stanford Unversty Stanford, CA

More information

Face Detection with Deep Learning

Face Detection with Deep Learning Face Detecton wth Deep Learnng Yu Shen Yus122@ucsd.edu A13227146 Kuan-We Chen kuc010@ucsd.edu A99045121 Yzhou Hao y3hao@ucsd.edu A98017773 Mn Hsuan Wu mhwu@ucsd.edu A92424998 Abstract The project here

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

Multi-view 3D Position Estimation of Sports Players

Multi-view 3D Position Estimation of Sports Players Mult-vew 3D Poston Estmaton of Sports Players Robbe Vos and Wlle Brnk Appled Mathematcs Department of Mathematcal Scences Unversty of Stellenbosch, South Afrca Emal: vosrobbe@gmal.com Abstract The problem

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

A mathematical programming approach to the analysis, design and scheduling of offshore oilfields

A mathematical programming approach to the analysis, design and scheduling of offshore oilfields 17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 A mathematcal programmng approach to the analyss, desgn and

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

Analysis of Continuous Beams in General

Analysis of Continuous Beams in General Analyss of Contnuous Beams n General Contnuous beams consdered here are prsmatc, rgdly connected to each beam segment and supported at varous ponts along the beam. onts are selected at ponts of support,

More information

Help for Time-Resolved Analysis TRI2 version 2.4 P Barber,

Help for Time-Resolved Analysis TRI2 version 2.4 P Barber, Help for Tme-Resolved Analyss TRI2 verson 2.4 P Barber, 22.01.10 Introducton Tme-resolved Analyss (TRA) becomes avalable under the processng menu once you have loaded and selected an mage that contans

More information

Hybrid Non-Blind Color Image Watermarking

Hybrid Non-Blind Color Image Watermarking Hybrd Non-Blnd Color Image Watermarkng Ms C.N.Sujatha 1, Dr. P. Satyanarayana 2 1 Assocate Professor, Dept. of ECE, SNIST, Yamnampet, Ghatkesar Hyderabad-501301, Telangana 2 Professor, Dept. of ECE, AITS,

More information

Detection of hand grasping an object from complex background based on machine learning co-occurrence of local image feature

Detection of hand grasping an object from complex background based on machine learning co-occurrence of local image feature Detecton of hand graspng an object from complex background based on machne learnng co-occurrence of local mage feature Shnya Moroka, Yasuhro Hramoto, Nobutaka Shmada, Tadash Matsuo, Yoshak Shra Rtsumekan

More information