Human Tracking in Thermal Catadioptric Omnidirectional Vision
|
|
- Julian Tucker
- 6 years ago
- Views:
Transcription
1 Proceedng of the IEEE Internatonal Conference on Informaton and Automaton Shenzhen, Chna June 20 Human Tracng n Thermal Catadoptrc Omndrectonal Vson Yazhe Tang, Youfu L, Tanxang Ba, Xaolong Zhou Department of Manufacturng Engneerng and Engneerng Management Cty Unversty of Hong Kong Kowloong, Hong Kong,Chna yztang2@student.ctyu.edu.h & meyfl@ctyu.edu.h Zhongwe L Department of Materal Scence and Engneerng Huazhong Unversty of Scence and Technology Wuhan, Hube Provnce, Chna zhongwl@ctyu.edu.h Abstract - We propose to explore a novel tracng system for human tracng n thermal catadoptrc omndrectonal (TCO) vson, whch s able to realze the survellance n all-weather and wde feld of vew condtons. In contrast, prevous human tracng system manly focuses on tracng n conventonal magng system. In ths paper, the proposed tracng method adopts the classfcaton posteror probablty of Support Vector Machne (SVM) to relate the observaton lelhood of partcle flter for effcent tracng. However, prevous wors only employ the fnal output label of SVM for classfcaton. Due to no exstng TCO vson dataset avalable n publc, we establsh a dataset ncludng TCO vdeos and extracted human samples to tran the classfer and test the proposed tracng method. Moreover, we adjust tracng wndow dstrbuton of partcle flter to ft the characterstc of catadoptrc omndrectonal vson whch s the sze of target n omn-mage depends on the dstance between target mage and the center of catadoptrc omndrectonal mage. Fnally, the expermental results show that our proposed tracng method has a stable and good performance n TCO vson tracng system. Index Terms Tracng, Thermal, Catadoptrc omndrectonal vson. I. INTRODUCTION Human tracng s a hot research topc n computer vson area, whch conssts of the man part of survellance system. In modern socety, the automatc survellance system becomes more and more popular n a large varety of applcatons. However, the prevous survellance system [] manly focuses on the conventonal vson system, whch has a lmted feld of vew and depends on the llumnaton. In ths paper, we try to explore a novel tracng system for human tracng n thermal catadoptrc omndrectonal (TCO) vson to realze survellance under all-weather and wde feld of vew condtons. The proposed novel survellance system can offer a much wder feld of vew and ndependent of llumnaton, whch permt to acqure more nformaton from the envronment. In order to overcome the llumnaton restrcton, the thermal camera s employed for t can detect the warmblooded anmal (Human) n any llumnaton condton, no matter day and nght, ran and fog. Furthermore, the prce of thermal camera has reduced dramatcally, so t becomes popular n cvlan applcaton over the recent years. To acheve a bg feld of vew, the catadoptrc omndrectonal sensor s able to meet our requrement, whch can reflect the surroundng lght (360 degree n horzon) nto a sngle camera. So, the hardware of proposed survellance system manly conssts of a thermal camera, a catadoptrc omndrectonal mrror and a stand (Fg. ). In addton, a representatve of TCO mage s gven n Fg. 2. For the merts of the proposed survellance system, t could have a wde range of applcatons, such as survellance n the arport, securty n the mltary, wld anmal conservaton and so on. Fg.. Confguraton of TCO system. Fg. 2. Representatve of TCO mage. Tracng n TCO vson has some bg challenges due to only lmted features can be used n thermal magery and severe nherent dstorton n catadoptrc omndrectonal vson. To the best of the author's nowledge, there are very lmted wors relate to detecton and tracng n TCO vson can be referenced by our wor. Prevously, researchers dd some excellent wors on detecton and tracng n thermal vson. Even tracng n thermal magng s dffcult compare to tradtonal vsble magng but drven by ts mert of llumnaton ndependent, and decreasng prce of t, more and more researchers began to focus on thermal magery vson. Recently, there have a flurry of wors on human detecton and tracng n thermal vson. [2] employed Support Vector Machne (SVM) for classfcaton and use of Kaman flter //$ IEEE 97
2 ntegrate mean shft to trac n thermal magery. In [3], the author presented a two-stage template-based method ntegrates wth an Adaboosted classfer for pedestran detecton. [4] ntegrated SVM [5][6] and Hstogram of Orented Gradent (HOG) [7] to detect the pedestran n thermal magery. In [8], the author used a generalzed expectaton-maxmzaton (EM) algorthm to separate nfrared mages nto bacground and foreground layers and ncorporate wth SVM for classfcaton, then present a graph matchng-based method to meet the tracng purpose. In ths decade, more researchers began to concentrate on omndrectonal vson tracng and detecton for ts wde feld of vew. In [9], a fsheye omndrectonal tracng system s ntroduced, the author used of optcal flow to detect the target and employed color feature based ernel partcle flter (KPF) to realze the sngle target tracng n omndrectonal vson. [0] presented a catadoptrc omndrectonal survellance system whch use of mult-bacground modelng and dynamc thresholdng to mae a outdoor tracng n clutter to spot the snper n the battlefeld. [][2] used of partcle flter ncorporates wth color feature to realze the tracng n catadoptrc omndrectonal vson. [3] ntroduced a catadoptrc omndrectonal pedestran recognton system for vehcle automaton, whch proposed a method of boosted cascade of wavelet-based classfers combned wth a subsequent texture-based neural networ. [3][4] unwarped the catadoptrc omndrectonal mage to the panoramc mage frst, and then mae the human recognton. However, I prefer to mae tracng n the orgnal catadoptrc omndrectonal mage rather than panoramc mage due to unwarp the omnmage nvolve nterpolaton whch s tme consumng. In addton, unwarppng has the rs to splt the target that locates at the border of the panoramc mage. In ths paper, we manly address the tracng n TCO vson. So, the proposed system s ntalzed manually n detecton phase. Due to only lmted features can be adopted n thermal magery, the HOG [7] feature wll be employed as descrptor for contour encodng. Le the most of tradtonal methods [4] [7] [5], our proposed method adopts SVM to classfy the extracted HOG feature n the TCO mage. More mportant s the proposed tracng method ntends to tae advantage of the SVM classfcaton posteror probablty to relate the observaton lelhood of partcle flter for effcent tracng n TCO vson (Fg. 3). For the reason of no exstng TCO vson dataset n publc, a dataset ncludng TCO vdeos and extracted human samples s formed. The extracted samples are used to tran the SVM classfer, and TCO vdeos are employed for testng. In addton, polar coordnate s adopted to ft the catadoptrc omndrectonal vson. Moreover, the tracng wndow dstrbuton of partcle flter s adjusted to ft the characterstc of catadoptrc omndrectonal vson whch s the sze of target n omn-mage depend on the dstance between target mage and center of catadoptrc omndrectonal mage. The paper s structured as follows. In Secton II, present the proposed tracng method whch conssts of partcle flter theoretcal foundaton, SVM classfcaton posteror probablty, extracted TCO samples and tracng wndow dstrbuton of partcle flter based on catadoptrc omndrectonal vson. Secton III presents and dscusses the expermental results. Fnally, the concluson wll be followed n Secton IV. Input Image Extracted HOG SVM Classfcaton Posteror Probablty Fg. 3. The framewor of proposed method Partcle Flter II. PROPOSED TRACKING ALGORITHM Tracng n TCO vson s dffcult due to there are several bg challenges have to be overcome, such as lmted features n thermal mage and nherent dstorton n catadoptrc omndrectonal vson. In ths secton, we propose to ntroduce a novel tracng method whch s very sutable to apply n TCO vson. The proposed method ntegrates SVM wth partcle flter and maes some relevant adjustments to ft the catadoptrc omndrectonal vson for effcent tracng. The detal of algorthm s shown as followng. A. Partcle Flter Partcle flter [6]-[8] s also nown as Sequental Monte Carlo methods (SMC), and t has been wdely used for ts advantage of nonlnear/non-gaussan. In Bayesan framewor, the partcle flter s recursvely obtan the state x at tme, gven the avalable observatons z z, z, : = 2, z up to tme p x z at tme - s avalable, the. Suppose posteror ( : ) posteror ( x z ) p : can be obtaned recursvely by predcton and update. The predcton stage maes use of the probablstc system transton model p ( x x ) to predct the posteror probablty of tme nstant. Predcton step: ( x z ) p( x x ) p( x z ) p dx () : = : At tme nstant, an observaton z s avalable, the posteror p ( x z : ) at tme nstant can be obtaned through update pror by Bayes s rule. Update step: ( x z ) : ( z x ) p( x z: ) p( z z ) p p = (2) : Where p ( z x ) s the observaton lelhood that nfluences the weght dstrbuton of each partcle. In ths paper, the observaton lelhood can be obtaned from the SVM classfcaton posteror probablty. At tme nstant, each partcle represents a hypothetcal state x, wth correspondng weght w. So, the prncple of partcle flter s to utlze fnte dscrete N weghted partcles 98
3 { x, w },2,, N = to approxmate the contnuous posteror dstrbuton. Of course, a large number of partcles can approach to the true dstrbuton wth a greatest degree. However, the computatonal cast s ncreased as the N ncreasng. So, that s a trade-off problem n the practcal stuaton. B. SVM Classfcaton Posteror Probablty SVM [9] [20] s a very useful algorthm for pattern classfcaton. It can dscrmnate the unlabeled samples correctly based on a traned model. Suppose the tranng sample set s provded as,, l s the number of the tranng samples. The SVM wors by maxmzng the margn between two classes n feature space as follows [20]: N T mn J( w, b, ξ) = w w + C ξ (3) 2 Subject to the constrants: = T y[ w ϕ( x ) + b] ξ, 0 Where s the slac varable, C s the penalty factor, ϕ( ) s the mappng from nput space to feature space. By tang the Lagrangan of (3), the orgnal problem can be derved as: l l max α αα yy(, ) x x (4) j j = 2 =, j= classfed by SVM, the result of (6) expresses the extent of extracted feature be postve (class ). That s p p =P(y = f) represents the probablty of sample x to be classfed nto postve set (class ). In contrast, p n =-P(y = f) s the probablty of negatve set (class -). So, the acheved SVM classfcaton posteror probablty can be employed to obtan observaton model p( z x = x ) as shown n the followng equaton: p 2 ( z x ) exp( d ) w w λ (7) ( z x ) p (8) Where s covarance and d s defned by (9). The obtaned observaton lelhood ( ) p z x s used to relate weght w of the partcles n (8). If the extracted sample s classfed nto postve set by SVM, the proposed algorthm calculates the dstance d. Oppostely, the algorthm gnores the result f the extracted sample s classfed nto negatve set. The obtaned d expresses the dstance between canddate sample and standard postve sample. In other word, f dstance d gettng small, the sample has a hgher probablty to be postve and the weght gettng to bg. (9) D. Extracted TCO Samples l = Subject to the constrants: α y = 0 α 0 Where ( x, x) = ϕ( x) ϕ( x ) s a ernel functon, and =,2, l, s the Lagrange coeffcents whch s obtaned from (4). The hyperplane functon can be expressed as: (a) Postve TCO samples; m f( x) = α y x x + b (5) = T Where m s the number of support vectors. Further, the obtaned hyperplane functon (5) can be used to derve the classfcaton posteror probablty [2] defned as follow: Py ( = f) = + exp (6) + ( Af B) It expresses the posteror probablty of ths sample be classfed nto class (postve set). Oppostely, -P(y= f) s the probablty of sample be classfed nto class - (negatve set). C. Lelhood As aforementoned, f an extracted local feature be (b) Negatve TCO samples; Fg. 4. Extracted TCO samples. In tradtonal detecton and tracng communty, some publc sample datasets are avalable. But for human tracng n TCO vson, there are no exstng datasets can be utlzed untl now. So, we have to manually extract the human (foreground) samples n TCO mages wth a tedous and tme consumng process. Also, the negatve samples are obtaned automatcally from a set of TCO mages not contanng human. Some extracted samples are shown n Fg. 4. The formed TCO dataset can be used to tran the classfer, test the proposed tracng method and further research on detecton and tracng n TCO. 99
4 E. Tracng Wndow Dstrbuton n TCO vson Tracng n TCO vson s dfferent wth tradtonal tracng system due to nherent dstorton caused by lght reflecton of convex catadoptrc mrror. So, the effectve omn-mage s the mage on catadoptrc mrror that s crcular. Naturally, the polar coordnate s employed for TCO tracng system, and the center of omn-mage s algnng wth the orgn of the polar coordnate. The orgn O (0, 0) of the mage XY coordnates s located on the top left corner. The center O (0, 0) of the polar coordnate can be obtaned (Fg. 5). Fg. 5. Polar coordnate n TCO tracng system. In catadoptrc omndrectonal vson system, the sze of target S n the mage vares wth the dstance D between target n mage and orgn of polar coordnate (Fg. 5). When the target approachng to the catadoptrc omndrectonal sensor from far to near n world coordnate, the sze of target S gettng to bg and dstance D gettng to small. That s to say, the sze of target S n mage depends on the dstance D. Based on forementoned characterstc of catadoptrc omndrectonal vson, the partcle flter can dstrbute the tracng wndow more reasonable. The sze of tracng wndow S s changng proportonally wth the dstance between partcle and center of omn-mage. In practcal stuaton, the real sze of tracng wndow s depend on the system confguraton, such as the curvature of catadoptrc mrror, the orentaton of catadoptrc mrror, the heght of catadoptrc mrror from ground and so on. In ths paper, we adopt HOG feature as the contour descrptor of target. We use a 2 by 2 cell array to form a bloc, each cell conssts of 4 pxels and the 9-bn hstogram of gradent magntude at each orentaton s computed. So, each bloc forms a 36-dmensonal vector. Fnally, the proposed tracng method s formed. The SVM classfcaton posteror probablty s ntegrated wth the partcle flter, and the tracng wndow dstrbuton based on characterstc of catadoptrc omndrectonal vson s proposed. Accordng to above, the proposed tracng method should able to trac the human n TCO vson effcently. The detal experments are show below. III. EXPERIMENT In the begnnng of ths paper, the confguraton of TCO system has been ntroduced. In ths secton, the experment results of proposed tracng method, gray-level nformaton based partcle flter and gray-level nformaton based onlne adaptve partcle flter are compared. Through test n a number of dfferent experments, the effectveness, stablty and robustness of proposed tracng algorthm are verfed. ) Proposed method s the proposed tracng method whch utlzes the SVM classfcaton posteror probablty to mpact the observaton lelhood of partcle flter. 2) G-PF s the partcle flter based on gray-level nformaton, whch has a fxed reference template. 3) A-PF s the onlne adaptve partcle flter based on gray-level nformaton, whch updates the reference template on tme nstead of fxed template. For a far comparson, we set the same parameters for above three tracng methods, such as partcles number N, covarance and so on. The dynamc models of the three tracng methods are random walng models whch represent as x = x + v, where v s a zero-mean Gaussan random varable. For Proposed method, the SVM classfer has been well traned wth 500 postve samples and 900 negatve samples. Due to lmted space, we just show three expermental results wth two shot n the daytme and one n the nghttme. In ths paper, t s should be noted that we just test sngle target to dscuss the feasblty of the Proposed method n TCO vson. After that, mult-targets tracng of TCO vson wll be extended n our future wor. In experment I, a person wals around the TCO-sensor wth a short tme occluson n the daytme. At the begnnng, three tracers are ntalzed respectvely, and they traced the target successfully (Fg. 6). After several frames later, A-PF tends to drft whch results n lost target at Frame 49 fnally (Fg. 7). Ths s not dffcult to understand snce A-PF s easy mpacted by the bacground although t good at adaptve to the changng of target. Moreover, only lmted gray-level nformaton can be adopted by A-PF n thermal mage nstead of three channels RBG features n normal vsble mage. In ths experment, only Proposed method and G-PF survve to trac the human target successfully to the end. Although both of A-PF and G-PF are based on gray-level nformaton, the reason of dfferent results s A-PF eeps update the template to adapt the change of target that easy tend to drft but G-PF eep the fxed template that s effectve when the dstracton from bacground s lmted. In addton, we set a short tme occluson to test reacton senstvty of the tracng methods when target recover from the occluson. As shown n the Fg. 8, the Proposed method can recover from the occluson n tme when target appear at Frame 332. In contrast, G-PF cannot capture the target sharply. Due to SVM classfer s well traned, the human target can be detected qucly even f the tracng wndow s not well center on the target. Instead, A-PF/G-PF requres relatve restrct template match. If the partcles cannot well land on the target, A-PF/G-PF has rs to fal. Especally n the stuaton of low number of partcles N (N=30 n our experment), the tracng performance of 00
5 Proposed method s much better than A-PF/G-PF n TCO vson. In experment II, a person wals from near to far relatve to the TCO-sensor n the daytme. Three tracers are well ntalzed as experment I and they trac the human target well n the frst few frames. Several frames later, A-PF begn to drft and lose the target completely at the Frame 03 due to accumulated changng of adaptve template (Fg. 9). In the meantme, Proposed method and G-PF stll wor on tracng well. After a few frames later, Proposed method and G-PF lose the target when the target dsappear at the dstant place for a whle. However, the Proposed method can capture the target mmedately when t appears agan because the partcles of Proposed method loo around the target at the nearby place where t dsappeared (Fg. 0). In contrast, G-PF loses the target forever due to the partcles of G-PF have drft to the wrong place where has smlar gray-level dstrbuton as target. Fg. 6. Experment I, Frame. From left to rght: Proposed method, G-PF, A-PF Fg. 7. Experment I, Frame 49. From left to rght: Proposed method, G-PF, A-PF Fg. 8. Experment I, Frame 332. From left to rght: Proposed method, G-PF, A-PF Fg. 9. Experment II, Frame 03. From left to rght: Proposed method, G-PF, A-PF Fg. 0. Experment II, Frame 202. From left to rght: Proposed method, G-PF, A-PF 0
6 Fg.. Experment III, Frame 32. From left to rght: Proposed method, G-PF, A-PF Fg. 2. Experment III, Frame 362. From left to rght: Proposed method, G-PF, A-PF In experment III, a person wals around the TCO-sensor n the nghttme, and system also well ntalzed at the begnnng. As prevous two experments, A-PF loses target agan at Frame 32 (Fg. ). At Frame 362, G-PF also loses the target completely wthout occluson and dsappearance but caused by bacground dstracton (Fg. 2). The reason of G- PF lost target s due to very smlar gray-level feature dstrbuton of surroundng bacground wth reference template at Frame 362. Fnally, only Proposed method tracs the human target well and survve tll the end of experment. In all experments, Proposed method not only tracs the human target successfully but well focus on center of target and very lmted bas occur. In contrast, G-PF and A-PF often happen to bas and drftng n some dffcult stuatons, whch cause to lose target n the end. In summary, the above three experments can effectve verfy the Proposed method has a good performance on tracng the human target n TCO vson. IV. CONCLUSION In ths paper, we present a novel human tracng system n TCO vson. The proposed tracng method maes an effcent utlzaton of SVM classfcaton posteror probablty to relate the observaton lelhood of partcle flter for tracng purpose. Moreover, the proposed tracng approach has acheved a good performance on tracng the human n TCO vson through a varety of experments. Fnally, we wll extend our research to mult-target tracng n TCO vson for real TCO survellance system n the future. REFERENCES [] P. Vola, M. J. Jones, and D. Snow. Detectng pedestrans usng patterns of moton and appearance. In Proc. ICCV, Volume 2, pages , 2003 [2] F. Xu, X. Lu, K. Fujmura, Pedestran detecton and tracng wth nght vson, IEEE Trans. Intell. Transport. Syst. 6 (2005) 63-7 [3] J. Davs, M. Kec, A two-stage approach to person detecton n thermal magery, Proc. Worshop on Applcatons of Computer Vson, 2005, IEEE OTCBVS WS Seres Bench [4] F. Suard, A. Raotomamonjy, A. Bensrhar, Pedestran detecton usng nfrared mages and hstograms of orented gradents, Intellgent Vehcles Symposum 2006, June 3-5, Toyo, Japan [5] C. J. C. Burges, A Tutoral on Support Vector Machnes for Pattern Recognton, Data Mnng and Knowledge Dscovery, 2, 2-67, 998 [6] V. Vapn. Statstcal Learnng Theory. Wley, 998 [7] N. Dalal, B. Trggs, Hstogram of Orented Gradents for Human Detecton, CVPR, volume 2, pp [8] C. X. Da, Y. F. Zheng, X. L, Layered representaton for pedestran detecton and tracng n nfrared magery, Computer Vson and Image Understandng. 06 (2007) [9] S. Y. Yang, G. W. Mn and C. Zhang, Tracng unnown movng targets on omndrectonal vson, Vson Research, 49 (2009), pp [0] T. E. Boult, X. Gao, R. Mcheals, M. Ecmann, Omn-drectonal vsual survellance, Image Vs. Comput. 22, (2004), pp [] J. C. Bazn, K. J. Yoon, I. Kweon, C Demonceaux, P Vasseur, Partcle flter approach adapted to catadoptrc mages for target tracng applcaton, IEEE, BMVC, [2] O. A Jame and B.C Eduardo, Omndrectonal vson tracng wth partcle flter, ICPR [3] W. Schulz, M. Enzwler and T. Ehlgen, Pedestran recognton from a movng catadoptrc camera, Proceedngs of the 29th DAGM conference on Pattern recognton 2007, pp [4] A. L. C Barcza, J. O. Jr and V. G. Jr, Face tracng usng a hyperbolc catadoptrc omndrectonal system, Res. Lett. Inf. Math. Sc, 2009 (3), pp55-67 [5] W. L. Ye, H. P. Lu, F. C. Sun and M. Gao, Vehcle tracng based on co-learnng partcle flter, IROS, 2009, pp [6] M. S. Arulampalam, S Masell, N Gordon and T Clapp, A Tutoral on partcle flters for onlne nonlnear/non-gaussan Bayesan tracng, IEEE Transactons on Sgnal Processng, (50), 2002, pp [7] A. Doucet, J.F.G. de Fretas and N.J.Gordon, An ntroducton to sequental Monte Carlo methods, n Sequental Monte Carlo Methods n Practce, A. Doucet, J.F.G.de Fretas and N.J.Gordon, Eds. New Yor: Sprnger-Verlag, 200 [8] M. Isard, A. Blae, Condensaton-condtonal desty propagaton for vsual tracng, IJCV, vol.29,no.,998, pp.5-28 [9] V. Vapn, The Nature of Statstcal Learnng Theory, Sprng-Verlag, New Yor, 995 [20] C. J. C. Burges, A Tutoral on Support Vector Machnes for Pattern Recognton, Data Mnng and Knowledge Dscovery, 2, 2-67, 998 [2] J. C. Platt. Probablstc outputs for support vector machnes and comparsons to regularzed lelhood methods. n Advances n Large Margn Classfers, Alexander J. Smola, Peter Bartlett, Bernhard Scholopf, Dale Schuurmans, eds., MIT Press, (999) 02
Outline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1
4/14/011 Outlne Dscrmnatve classfers for mage recognton Wednesday, Aprl 13 Krsten Grauman UT-Austn Last tme: wndow-based generc obect detecton basc ppelne face detecton wth boostng as case study Today:
More informationThe 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 informationAPPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET
APPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET Jae-young Lee, Shahram Payandeh, and Ljljana Trajovć School of Engneerng Scence Smon Fraser Unversty 8888 Unversty
More informationEYE CENTER LOCALIZATION ON A FACIAL IMAGE BASED ON MULTI-BLOCK LOCAL BINARY PATTERNS
P.G. Demdov Yaroslavl State Unversty Anatoly Ntn, Vladmr Khryashchev, Olga Stepanova, Igor Kostern EYE CENTER LOCALIZATION ON A FACIAL IMAGE BASED ON MULTI-BLOCK LOCAL BINARY PATTERNS Yaroslavl, 2015 Eye
More informationAPPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET
APPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET Jae-young Lee, Shahram Payandeh, and Ljljana Trajovć School of Engneerng Scence Smon Fraser Unversty 8888 Unversty
More informationObject Contour Tracking Using Multi-feature Fusion based Particle Filter
Object Contour Tracng Usng Mult-feature Fuson based Partcle Flter Xaofeng Lu 1,3, L Song 1,2, Songyu Yu 1, Nam Lng 2 Insttute of Image Communcaton and Informaton Processng 1 Shangha Jao Tong Unversty,
More informationLearning 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 informationA 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 informationSupport Vector Machines
/9/207 MIST.6060 Busness Intellgence and Data Mnng What are Support Vector Machnes? Support Vector Machnes Support Vector Machnes (SVMs) are supervsed learnng technques that analyze data and recognze patterns.
More informationImprovement 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 informationHistogram of Template for Pedestrian Detection
PAPER IEICE TRANS. FUNDAMENTALS/COMMUN./ELECTRON./INF. & SYST., VOL. E85-A/B/C/D, No. xx JANUARY 20xx Hstogram of Template for Pedestran Detecton Shaopeng Tang, Non Member, Satosh Goto Fellow Summary In
More informationCluster 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 informationRobust Inlier Feature Tracking Method for Multiple Pedestrian Tracking
2011 Internatonal Conference on Informaton and Intellgent Computng IPCSIT vol.18 (2011) (2011) IACSIT Press, Sngapore Robust Inler Feature Trackng Method for Multple Pedestran Trackng Young-Chul Lm a*
More informationContent 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 informationSupport Vector Machines
Support Vector Machnes Decson surface s a hyperplane (lne n 2D) n feature space (smlar to the Perceptron) Arguably, the most mportant recent dscovery n machne learnng In a nutshell: map the data to a predetermned
More informationTracking by Cluster Analysis of Feature Points and Multiple Particle Filters 1
Tracng by Cluster Analyss of Feature Ponts and Multple Partcle Flters 1 We Du, Justus Pater Unversty of Lège, Department of Electrcal Engneerng and Computer Scence, Insttut Montefore, B28, Sart Tlman Campus,
More informationParallelism 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 informationOnline Detection and Classification of Moving Objects Using Progressively Improving Detectors
Onlne Detecton and Classfcaton of Movng Objects Usng Progressvely Improvng Detectors Omar Javed Saad Al Mubarak Shah Computer Vson Lab School of Computer Scence Unversty of Central Florda Orlando, FL 32816
More informationDiscriminative classifiers for object classification. Last time
Dscrmnatve classfers for object classfcaton Thursday, Nov 12 Krsten Grauman UT Austn Last tme Supervsed classfcaton Loss and rsk, kbayes rule Skn color detecton example Sldng ndo detecton Classfers, boostng
More informationClassifier Swarms for Human Detection in Infrared Imagery
Classfer Swarms for Human Detecton n Infrared Imagery Yur Owechko, Swarup Medasan, and Narayan Srnvasa HRL Laboratores, LLC 3011 Malbu Canyon Road, Malbu, CA 90265 {owechko, smedasan, nsrnvasa}@hrl.com
More informationEdge 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 informationA 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 informationLocal Quaternary Patterns and Feature Local Quaternary Patterns
Local Quaternary Patterns and Feature Local Quaternary Patterns Jayu Gu and Chengjun Lu The Department of Computer Scence, New Jersey Insttute of Technology, Newark, NJ 0102, USA Abstract - Ths paper presents
More informationBOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET
1 BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET TZU-CHENG CHUANG School of Electrcal and Computer Engneerng, Purdue Unversty, West Lafayette, Indana 47907 SAUL B. GELFAND School
More informationSLAM 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 informationSkew 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 informationIndex Terms Object tracking, Extended Kalmanfiter, Particle filter, Color matching.
Object Tracng under Heavy Occluson based on Extended Kalman Flter, Partcle Flter, and Color Matchng Youngsung Soh, Mudasar Qadr, Had Raja, Yongsu Hae, Intae Km, Mal M. Khan, and Tayyab Wahab soh@mju.ac.r,
More informationFast Feature Value Searching for Face Detection
Vol., No. 2 Computer and Informaton Scence Fast Feature Value Searchng for Face Detecton Yunyang Yan Department of Computer Engneerng Huayn Insttute of Technology Hua an 22300, Chna E-mal: areyyyke@63.com
More informationReal-time Joint Tracking of a Hand Manipulating an Object from RGB-D Input
Real-tme Jont Tracng of a Hand Manpulatng an Object from RGB-D Input Srnath Srdhar 1 Franzsa Mueller 1 Mchael Zollhöfer 1 Dan Casas 1 Antt Oulasvrta 2 Chrstan Theobalt 1 1 Max Planc Insttute for Informatcs
More informationA 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 informationFEATURE 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 informationWhat is Object Detection? Face Detection using AdaBoost. Detection as Classification. Principle of Boosting (Schapire 90)
CIS 5543 Coputer Vson Object Detecton What s Object Detecton? Locate an object n an nput age Habn Lng Extensons Vola & Jones, 2004 Dalal & Trggs, 2005 one or ultple objects Object segentaton Object detecton
More informationDetection 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 informationWishing 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 informationMULTISPECTRAL 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 informationKernel Collaborative Representation Classification Based on Adaptive Dictionary Learning
Internatonal Journal of Intellgent Informaton Systems 2018; 7(2): 15-22 http://www.scencepublshnggroup.com/j/js do: 10.11648/j.js.20180702.11 ISSN: 2328-7675 (Prnt); ISSN: 2328-7683 (Onlne) Kernel Collaboratve
More informationA novel framework for motion segmentation and tracking by clustering incomplete trajectories
A novel framewor for moton segmentaton and tracng by clusterng ncomplete trajectores Vasleos Karavasls, Konstantnos Bleas, Chrstophoros Nou Department of Computer Scence, Unversty of Ioannna, PO Box 1186,
More informationA Background Subtraction for a Vision-based User Interface *
A Background Subtracton for a Vson-based User Interface * Dongpyo Hong and Woontack Woo KJIST U-VR Lab. {dhon wwoo}@kjst.ac.kr Abstract In ths paper, we propose a robust and effcent background subtracton
More informationA New Feature of Uniformity of Image Texture Directions Coinciding with the Human Eyes Perception 1
A New Feature of Unformty of Image Texture Drectons Concdng wth the Human Eyes Percepton Xng-Jan He, De-Shuang Huang, Yue Zhang, Tat-Mng Lo 2, and Mchael R. Lyu 3 Intellgent Computng Lab, Insttute of Intellgent
More informationAn efficient method to build panoramic image mosaics
An effcent method to buld panoramc mage mosacs Pattern Recognton Letters vol. 4 003 Dae-Hyun Km Yong-In Yoon Jong-Soo Cho School of Electrcal Engneerng and Computer Scence Kyungpook Natonal Unv. Abstract
More informationClassifying Acoustic Transient Signals Using Artificial Intelligence
Classfyng Acoustc Transent Sgnals Usng Artfcal Intellgence Steve Sutton, Unversty of North Carolna At Wlmngton (suttons@charter.net) Greg Huff, Unversty of North Carolna At Wlmngton (jgh7476@uncwl.edu)
More informationCollaboratively Regularized Nearest Points for Set Based Recognition
Academc Center for Computng and Meda Studes, Kyoto Unversty Collaboratvely Regularzed Nearest Ponts for Set Based Recognton Yang Wu, Mchhko Mnoh, Masayuk Mukunok Kyoto Unversty 9/1/013 BMVC 013 @ Brstol,
More informationAn 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 informationRobust visual tracking based on Informative random fern
5th Internatonal Conference on Computer Scences and Automaton Engneerng (ICCSAE 205) Robust vsual trackng based on Informatve random fern Hao Dong, a, Ru Wang, b School of Instrumentaton Scence and Opto-electroncs
More informationThe Comparison of Calibration Method of Binocular Stereo Vision System Ke Zhang a *, Zhao Gao b
3rd Internatonal Conference on Materal, Mechancal and Manufacturng Engneerng (IC3ME 2015) The Comparson of Calbraton Method of Bnocular Stereo Vson System Ke Zhang a *, Zhao Gao b College of Engneerng,
More informationFace Recognition using 3D Directional Corner Points
2014 22nd Internatonal Conference on Pattern Recognton Face Recognton usng 3D Drectonal Corner Ponts Xun Yu, Yongsheng Gao School of Engneerng Grffth Unversty Nathan, QLD, Australa xun.yu@grffthun.edu.au,
More informationCS246: Mining Massive Datasets Jure Leskovec, Stanford University
CS46: Mnng Massve Datasets Jure Leskovec, Stanford Unversty http://cs46.stanford.edu /19/013 Jure Leskovec, Stanford CS46: Mnng Massve Datasets, http://cs46.stanford.edu Perceptron: y = sgn( x Ho to fnd
More informationMULTISPECTRAL REMOTE SENSING IMAGE CLASSIFICATION WITH MULTIPLE FEATURES
MULISPECRAL REMOE SESIG IMAGE CLASSIFICAIO WIH MULIPLE FEAURES QIA YI, PIG GUO, Image Processng and Pattern Recognton Laboratory, Bejng ormal Unversty, Bejng 00875, Chna School of Computer Scence and echnology,
More informationA Probabilistic Approach to Detect Urban Regions from Remotely Sensed Images Based on Combination of Local Features
A Probablstc Approach to Detect Urban Regons from Remotely Sensed Images Based on Combnaton of Local Features Berl Sırmaçek German Aerospace Center (DLR) Remote Sensng Technology Insttute Weßlng, 82234,
More informationIncremental Multiple Kernel Learning for Object Recognition
Incremental Multple Kernel Learnng for Obect Recognton Anruddha Kembhav, Behat Sddque, Roland Mezano, Scott McClosey, Larry S. Davs Unversty of Maryland, College Par Honeywell Labs Abstract A good tranng
More informationFace Recognition Method Based on Within-class Clustering SVM
Face Recognton Method Based on Wthn-class Clusterng SVM Yan Wu, Xao Yao and Yng Xa Department of Computer Scence and Engneerng Tong Unversty Shangha, Chna Abstract - A face recognton method based on Wthn-class
More informationBAYESIAN MULTI-SOURCE DOMAIN ADAPTATION
BAYESIAN MULTI-SOURCE DOMAIN ADAPTATION SHI-LIANG SUN, HONG-LEI SHI Department of Computer Scence and Technology, East Chna Normal Unversty 500 Dongchuan Road, Shangha 200241, P. R. Chna E-MAIL: slsun@cs.ecnu.edu.cn,
More informationAn 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 informationDiscriminative Dictionary Learning with Pairwise Constraints
Dscrmnatve Dctonary Learnng wth Parwse Constrants Humn Guo Zhuoln Jang LARRY S. DAVIS UNIVERSITY OF MARYLAND Nov. 6 th, Outlne Introducton/motvaton Dctonary Learnng Dscrmnatve Dctonary Learnng wth Parwse
More informationMargin-Constrained Multiple Kernel Learning Based Multi-Modal Fusion for Affect Recognition
Margn-Constraned Multple Kernel Learnng Based Mult-Modal Fuson for Affect Recognton Shzh Chen and Yngl Tan Electrcal Engneerng epartment The Cty College of New Yor New Yor, NY USA {schen, ytan}@ccny.cuny.edu
More informationLecture 13: High-dimensional Images
Lec : Hgh-dmensonal Images Grayscale Images Lecture : Hgh-dmensonal Images Math 90 Prof. Todd Wttman The Ctadel A grayscale mage s an nteger-valued D matrx. An 8-bt mage takes on values between 0 and 55.
More informationAdaptive Silhouette Extraction and Human Tracking in Dynamic. Environments 1
Adaptve Slhouette Extracton and Human Trackng n Dynamc Envronments 1 X Chen, Zhha He, Derek Anderson, James Keller, and Marjore Skubc Department of Electrcal and Computer Engneerng Unversty of Mssour,
More information12/2/2009. Announcements. Parametric / Non-parametric. Case-Based Reasoning. Nearest-Neighbor on Images. Nearest-Neighbor Classification
Introducton to Artfcal Intellgence V22.0472-001 Fall 2009 Lecture 24: Nearest-Neghbors & Support Vector Machnes Rob Fergus Dept of Computer Scence, Courant Insttute, NYU Sldes from Danel Yeung, John DeNero
More informationFace Recognition Based on SVM and 2DPCA
Vol. 4, o. 3, September, 2011 Face Recognton Based on SVM and 2DPCA Tha Hoang Le, Len Bu Faculty of Informaton Technology, HCMC Unversty of Scence Faculty of Informaton Scences and Engneerng, Unversty
More informationA 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 informationCHAPTER 3 SEQUENTIAL MINIMAL OPTIMIZATION TRAINED SUPPORT VECTOR CLASSIFIER FOR CANCER PREDICTION
48 CHAPTER 3 SEQUENTIAL MINIMAL OPTIMIZATION TRAINED SUPPORT VECTOR CLASSIFIER FOR CANCER PREDICTION 3.1 INTRODUCTION The raw mcroarray data s bascally an mage wth dfferent colors ndcatng hybrdzaton (Xue
More informationDynamic 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 informationImplementation of Robust HOG-SVM based Pedestrian Classification
Implementaton of Robust HOG-SVM based Pedestran Classfcaton Reecha P. Yadav K.K.W.I.E.E.R Nashk Inda Vnuchackravarthy Senthamlarasu and Krshnan Kutty KPIT Technologes Ltd. Pune Inda Sunta P. Ugale K.K.W.I.E.E.R
More informationAvailable online at Available online at Advanced in Control Engineering and Information Science
Avalable onlne at wwwscencedrectcom Avalable onlne at wwwscencedrectcom Proceda Proceda Engneerng Engneerng 00 (2011) 15000 000 (2011) 1642 1646 Proceda Engneerng wwwelsevercom/locate/proceda Advanced
More informationLearning a Class-Specific Dictionary for Facial Expression Recognition
BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 16, No 4 Sofa 016 Prnt ISSN: 1311-970; Onlne ISSN: 1314-4081 DOI: 10.1515/cat-016-0067 Learnng a Class-Specfc Dctonary for
More informationSVM Based Forest Fire Detection Using Static and Dynamic Features
DOI: 10.2298/CSIS101012030Z SVM Based Forest Fre Detecton Usng Statc and Dynamc Features Janhu Zhao, Zhong Zhang, Shzhong Han, Chengzhang Qu Zhyong Yuan, and Dengy Zhang Computer School, Wuhan Unversty,
More informationCorner-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 informationA high precision collaborative vision measurement of gear chamfering profile
Internatonal Conference on Advances n Mechancal Engneerng and Industral Informatcs (AMEII 05) A hgh precson collaboratve vson measurement of gear chamferng profle Conglng Zhou, a, Zengpu Xu, b, Chunmng
More informationLecture 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 informationAnytime Predictive Navigation of an Autonomous Robot
Anytme Predctve Navgaton of an Autonomous Robot Shu Yun Chung Department of Mechancal Engneerng Natonal Tawan Unversty Tape, Tawan Emal shuyun@robot0.me.ntu.edu.tw Abstract To acheve fully autonomous moble
More informationAUTOMATIC RECOGNITION OF TRAFFIC SIGNS IN NATURAL SCENE IMAGE BASED ON CENTRAL PROJECTION TRANSFORMATION
AUTOMATIC RECOGNITION OF TRAFFIC SIGNS IN NATURAL SCENE IMAGE BASED ON CENTRAL PROJECTION TRANSFORMATION Ka Zhang a, Yehua Sheng a, Pefang Wang b, Ln Luo c, Chun Ye a, Zhjun Gong d a Key Laboratory of
More informationTsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance
Tsnghua Unversty at TAC 2009: Summarzng Mult-documents by Informaton Dstance Chong Long, Mnle Huang, Xaoyan Zhu State Key Laboratory of Intellgent Technology and Systems, Tsnghua Natonal Laboratory for
More information3D vector computer graphics
3D vector computer graphcs Paolo Varagnolo: freelance engneer Padova Aprl 2016 Prvate Practce ----------------------------------- 1. Introducton Vector 3D model representaton n computer graphcs requres
More informationA 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 informationMetrol. Meas. Syst., Vol. XXIII (2016), No. 1, pp METROLOGY AND MEASUREMENT SYSTEMS. Index , ISSN
Metrol. Meas. Syst., Vol. XXIII (2016), No. 1, pp. 27 36. METROLOGY AND MEASUREMENT SYSTEMS Index 330930, ISSN 0860-8229 www.metrology.pg.gda.pl HISTOGRAM OF ORIENTED GRADIENTS WITH CELL AVERAGE BRIGHTNESS
More informationCS 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 informationGender Classification using Interlaced Derivative Patterns
Gender Classfcaton usng Interlaced Dervatve Patterns Author Shobernejad, Ameneh, Gao, Yongsheng Publshed 2 Conference Ttle Proceedngs of the 2th Internatonal Conference on Pattern Recognton (ICPR 2) DOI
More informationThe Study of Remote Sensing Image Classification Based on Support Vector Machine
Sensors & Transducers 03 by IFSA http://www.sensorsportal.com The Study of Remote Sensng Image Classfcaton Based on Support Vector Machne, ZHANG Jan-Hua Key Research Insttute of Yellow Rver Cvlzaton and
More informationFuzzy 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 informationVol. 5, No. 3 March 2014 ISSN Journal of Emerging Trends in Computing and Information Sciences CIS Journal. All rights reserved.
Journal of Emergng Trends n Computng and Informaton Scences 009-03 CIS Journal. All rghts reserved. http://www.csjournal.org Unhealthy Detecton n Lvestock Texture Images usng Subsampled Contourlet Transform
More informationA 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 informationReal-time Multiple Objects Tracking with Occlusion Handling in Dynamic Scenes
Real-tme Multple Obects Tracng wth Occluson Handlng n Dynamc Scenes Tao Yang 1, Stan Z.L 2, Quan Pan 1, Jng L 1 1 College of Automatc Control, Northwestern Polytechncal Unversty, X an, Chna, 710072 2 Natonal
More informationA 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 informationPCA Based Gait Segmentation
Honggu L, Cupng Sh & Xngguo L PCA Based Gat Segmentaton PCA Based Gat Segmentaton Honggu L, Cupng Sh, and Xngguo L 2 Electronc Department, Physcs College, Yangzhou Unversty, 225002 Yangzhou, Chna 2 Department
More informationTracking and Guiding Multiple Laser Beams for Beam and Target Alignment
Internatonal Journal of Automaton and Computng 12(6), December 2015, 600-610 DOI: 10.1007/s11633-015-0908-8 Trackng and Gudng Multple Laser Beams for Beam and Target Algnment Peng-Cheng Zhang De Xu Research
More informationNetwork Intrusion Detection Based on PSO-SVM
TELKOMNIKA Indonesan Journal of Electrcal Engneerng Vol.1, No., February 014, pp. 150 ~ 1508 DOI: http://dx.do.org/10.11591/telkomnka.v1.386 150 Network Intruson Detecton Based on PSO-SVM Changsheng Xang*
More informationLearning-based License Plate Detection on Edge Features
Learnng-based Lcense Plate Detecton on Edge Features Wng Teng Ho, Woo Hen Yap, Yong Haur Tay Computer Vson and Intellgent Systems (CVIS) Group Unverst Tunku Abdul Rahman, Malaysa wngteng_h@yahoo.com, woohen@yahoo.com,
More informationDETECTION OF MOVING OBJECT BY FUSION OF COLOR AND DEPTH INFORMATION
INTERNATIONAL JOURNAL ON SMART SENSING AN INTELLIGENT SYSTEMS VOL. 9, NO., MARCH 206 ETECTION OF MOVING OBJECT BY FUSION OF COLOR AN EPTH INFORMATION T. T. Zhang,G. P. Zhao and L. J. Lu School of Automaton
More informationRobust Mean Shift Tracking with Corrected Background-Weighted Histogram
Robust Mean Shft Trackng wth Corrected Background-Weghted Hstogram Jfeng Nng, Le Zhang, Davd Zhang and Chengke Wu Abstract: The background-weghted hstogram (BWH) algorthm proposed n [] attempts to reduce
More informationLine-based Camera Movement Estimation by Using Parallel Lines in Omnidirectional Video
01 IEEE Internatonal Conference on Robotcs and Automaton RverCentre, Sant Paul, Mnnesota, USA May 14-18, 01 Lne-based Camera Movement Estmaton by Usng Parallel Lnes n Omndrectonal Vdeo Ryosuke kawansh,
More informationFast Sparse Gaussian Processes Learning for Man-Made Structure Classification
Fast Sparse Gaussan Processes Learnng for Man-Made Structure Classfcaton Hang Zhou Insttute for Vson Systems Engneerng, Dept Elec. & Comp. Syst. Eng. PO Box 35, Monash Unversty, Clayton, VIC 3800, Australa
More informationABSTRACT 1. INTRODUCTION
Arborne Target Trackng Algorthm aganst Oppressve Decoys n Infrared Imagery Xechang Sun, Tanxu Zhang State Key Laboratory for Multspectral Informaton Processng Technologes; Insttute for Pattern Recognton
More informationCategorizing objects: of appearance
Categorzng objects: global and part-based models of appearance UT Austn Generc categorzaton problem 1 Challenges: robustness Realstc scenes are crowded, cluttered, have overlappng objects. Generc category
More informationFace 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 informationA 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 informationDetermining 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 informationAn Entropy-Based Approach to Integrated Information Needs Assessment
Dstrbuton Statement A: Approved for publc release; dstrbuton s unlmted. An Entropy-Based Approach to ntegrated nformaton Needs Assessment June 8, 2004 Wllam J. Farrell Lockheed Martn Advanced Technology
More informationUAV global pose estimation by matching forward-looking aerial images with satellite images
The 2009 IEEE/RSJ Internatonal Conference on Intellgent Robots and Systems October -5, 2009 St. Lous, USA UAV global pose estmaton by matchng forward-lookng aeral mages wth satellte mages Kl-Ho Son, Youngbae
More informationUser 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 informationSequential Monte-Carlo Based Road Region Segmentation Algorithm with Uniform Spatial Sampling
IPSJ Transactons on Computer Vson and Applcatons Vol.8 1 10 (Feb. 2016) [DOI: 10.2197/psjtcva.8.1] Regular Paper Sequental Monte-Carlo Based Road Regon Segmentaton Algorthm wth Unform Spatal Samplng Zdeněk
More information